Tuesday, December 29, 2015

A Lander for NASA’s Europa Mission

“This Act includes $1,631,000,000 for Planetary Science.  Of this amount, $261,000,000 is for Outer Planets, of which $175,000,000 is for the Jupiter Europa clipper mission and clarifies that this mission shall include an orbiter with a lander that will include competitively selected instruments and that funds shall be used to finalize the mission design concept with a target launch date of 2022.”

“…$175,000,000 is for an orbiter with a lander to meet the science goals for the Jupiter Europa mission as outlined in the most recent planetary science decadal survey.  That the National Aeronautics and Space Administration shall use the Space Launch System as the launch vehicle for the Jupiter Europa mission, plan for a launch no later than 2022, and include in the fiscal year 2017 budget the 5-year funding profile necessary to achieve these goals.”

- Final budget law for Fiscal Year 2016 regarding NASA’s Europa mission

While there’s at least eight years until it launches, this has been a pivotal year for developing NASA’s Europa mission.  Last spring, NASA selected a rich and highly capable instrument set.  This summer, following a design concept review, the mission moved from concept studies to an official mission.  And just last week, Congress directed NASA to expand the mission by adding a small lander as well as launch the mission by 2022 and use the Space Launch System. These latter aren’t just suggestions: they are the law.

There’s been almost no official information on the lander.  What we know comes from a long article from Ars Technica’s Eric Berger on the then possible addition of a lander and a dedicated plume flyby sub-satellite.   Berger is a long time space reporter and has developed a good relationship with House Appropriations Subcommittee Chairman John Culberson (R-TX).  (I make sure I read all of Berger’s articles.)  As Berger describes in detail in his article, Culberson has been the driving force behind the aggressive funding for this mission. 

In addition to an earlier launch, .Culberson also has wanted to see the mission carry a lander in addition to the mother craft that would make at least 45 close flybys of the moon.  In prior years, Culberson added funding to NASA’s budget specifically to study a lander option, and the Jet Propulsion Laboratory’s engineers have been studying options.  Berger’s story is focused more on Culberson, but it does provide a number of facts about the possible design for the lander:

The leading concept for the lander would be a small lander, perhaps about 230 kg with 20-30 kg for instruments.  For comparison, the 1996 Mars Pathfinder lander had a mass of 265 kg.

The lander would be delivered to Jovian orbit by the main spacecraft and then released in a high parking orbit well outside the intense radiation fields at Europa’s orbit.  The main spacecraft would study Europa’s surface for two to three years during its flybys to find the best combination of a scientifically interesting and safe landing spot.  

The actual landing would use the same skycrane approach used by the Curiosity Mars mission to deliver the lander safely to the surface. 
The lander would likely last perhaps 10 days on the surface using battery power.  During the lander’s lifetime, it would investigate the chemistry of the surface using a mass spectrometer and possibly a Raman spectrometer.

A lander could add $700M or more to the mission cost.  The last cost estimate I heard for just the main spacecraft was $2.1B.  We don’t know how firm the lander cost estimate is.

Adding a lander would delay launch from a possible 2022 to 2023.

This description is pretty bare bones, but with a little legwork, it is possible to flesh out these ideas with some informed speculation.  It helps that a number of previous studies have been published that examined concepts for a Europa lander.

In the mid-2000s, NASA studied a small Europa lander that would have had similar mass and capabilities to those reportedly be considered for the approved Europa mission. Credit: NASA/JPL.

The primary goal of any lander would be to sample material from the interior ocean to see if the chemicals needed to support life are present and whether complex organic molecules suggesting biotic or pre-biotic activity exist.  We lack the technology to drill through the kilometers of ice to reach the ocean directly.  However, in many locations the icy shell appears to be fractured and water from below has spilled onto the surface and frozen and in certain locations may be actively venting into space.  The goal will be to set the lander down in one of these zones.

Our current knowledge of Europa’s surface is too poor to select the scientifically most interesting sites that are also safe to land in.  The main spacecraft will spend three years circling Jupiter and flying low over Europa 45 times.  One of its prime goals will be to for its cameras and spectrometers to find the optimal combination of evidence of ocean material on the surface with a safe landing zone.  Any landing will need to wait for scientists to build their high resolution maps.

One aspect of this proposed lander concept is different than those I’ve seen before.  Most lander studies have looked at small spacecraft (and this proposal would count as a small spacecraft) that would be carried by the mother craft until just before landing.  For the design Berger reported on, lander and its descent stage would orbit Jupiter on their own for months to years before landing.  This means that together they are a fully functional independent spacecraft with its own solar arrays for power, propulsion, navigation, and communications.  Apparently the cost and mass of adding these functions to the descent stage and lander is a better bargain than adding the radiation hardening that would be required if the lander were carried past Europa 45 times.

Once on the surface, the lander could be well protected from radiation.  The rotation of Jupiter’s magnetosphere causes the radiation to slam into Europa’s trailing hemisphere.  The leading hemisphere has Europa’s bulk as a very effective radiation shield, and radiation there is fairly low.  Past proposals have focused on putting a lander on the leading hemisphere.  As a result, the lander likely would run out of power before radiation would fry its electronics.  Fortunately, there are several regions on the leading hemisphere where the icy shell appears to have been recently (in geologic terms, anyway) fractured.

Berger’s article states that the lander likely would be powered by batteries, limiting its life to around 10 days.  Solar panels apparently are being considered, but I can see why they might not be attractive.  Sunlight at Jupiter is weak, and solar panels large enough to harvest a meaningful amount of that light might be too bulky and heavy for the mission. 

Berger’s article lists just two possible instruments for the lander.  Based on his language, the core instrument would be a mass spectrometer that would “weigh” the molecules and atoms in samples scooped, cut, or drilled from the surface.  Extremely complex molecules could suggest life, especially if they are rich in elements, like carbon, which are the basis for life on the Earth.  A second instrument under consideration would be a Raman spectrometer that would illuminate samples with lasers and use the resulting “glow” to measure composition including complex organic molecules.  (For those who understand Raman spectroscopy, please forgive this simplification of a complex subject; here’s a link to a Wikipedia article for more on this technology.)  I’ve also heard through other sources that the lander would carry an imager to examine the terrain around the landing site.

Once on the surface, the lander would use a sample acquisition system to collect a sample of ice from the surface.  As Berger points out, at Europa’s surface temperatures, the ice is as hard as rock, so the cutting or drilling mechanism will need to be robust.  After the sample is collected, it would be delivered to the instruments to measure its composition.  If the lander touches down near an active vent, the mass spectrometer could also measure the composition of the particles and gases in the plume.

Previous studies typically have proposed at least two other instruments.  Europa’s icy shell is constantly being stressed by the tides induced by Jupiter which should produce high seismic activity.  A seismometer would give scientists a rich data set on the interior structure of the ice.  Europa also sits within Jupiter’s intense magnetosphere which causes an induced magnetic field in the moon’s interior ocean.  How this induced field varies as Europa orbits Jupiter would provide valuable clues to the size and salinity of the ocean.  A magnetometer on the lander could provide continuous measurements for the life of the lander.  Berger’s article was silent on whether or not these instruments are under consideration for this version of the lander. 

(On a side note, a magnetometer plus a simple plasma probe would allow the lander to conduct useful science while it orbits Jupiter waiting for landing.  Scientists would like to study Jupiter’s magnetosphere from multiple locations at once.  The lander while in orbit around Jupiter could complement similar measurements from the main spacecraft, and depending on the timing, also from Europe’s JUICE spacecraft that will enter Jovian orbit in the late 2020s.)

Berger’s article is silent on how data would be returned to Earth.  Two possibilities are obvious – low data rate transmissions directly from the lander to Earth or high data rate transmissions from the lander to the orbiter for later relay to Earth.  Data relay from the mother flyby spacecraft likely would be possible, but the rapidly changing relative locations of the landing site and the orbiter circling Jupiter may limit how much data could be returned and when communication relay is possible.  A recent European study for a Europa lander assumed that the mother spacecraft would have just one chance to directly receive data from the lander in a 10 day period.  One argument for excluding a seismometer is that this instrument would produce large amounts of data that may be difficult to return directly to Earth.  The European study found that the brief relay between lander and orbiter would have enabled the return of seismic data.   Magnetometers, on the other hand, produce only small amounts of data that likely could be directly relayed to Earth (assuming the lander would have that ability). 

In the highest resolution images obtained by the Galileo Jupiter orbiter, the regions of Europa that appear to be fractured and have possible ocean material on the surface are extremely rugged; these cliffs are approximately 10 stories tall.  Credit: NASA/JPL.

A major challenge for any Europa lander will be that the scientifically most interesting places to study also appear to have extremely rugged terrain which makes landing risky.  Berger’s article briefly mentions that the lander would use an autonomous landing system to examine the terrain below it to pick out safe spots to put down.  Technologies to allow a lander to image its landing site during descent have been studied for years and were implemented on the Chinese Chang’e 3 lunar lander and are under consideration for NASA’s 2020 Mars rover.   During final descent, these systems use images taken by the lander in real time to analyze the terrain below to identify safe landing zones.  With an autonomous guided descent, scientists can target an area that overall is rugged but has small safe zones.

What I conclude from the clues Berger supplies is that the Europa landing would be much like the Philae comet lander (although with Europa’s higher gravity, the lander will not bounce across the surface after touchdown as Philae did).  The lander would have just a few days to conduct its operations and return the data to Earth.  On Mars, we have become accustomed to landed missions that last years with plenty of time to carefully consider where to sample and then conduct follow up studies.  A Europa lander will be a mad dash to complete the science goals before the batteries die.

By the end of its life, the lander will have returned our first data directly from the surface of an ocean world that may harbor life that arose independently from the life on Earth.

Launch on the Space Launch System (SLS) booster currently being developed by NASA primarily to support human exploration would significantly shorten the flight time to Jupiter.  Other presentations list the nominal cost for an Atlas launch at ~$350M and for an SLS launch at ~$500M.  Reducing the cost of operations during the flight to Europa could make the SLS option, which is currently required by law, equal to or cheaper than the Atlas option.  Credit: NASA/JPL.
Editorial Thoughts:  I of course want to see a lander delivered to the surface of Europa, but I have mixed feelings about the inclusion of a lander on NASA’s first dedicated mission to Europa for two reasons.  First, as I will explore in more detail in my next post, adding a lander to the existing Europa mission will push its costs up, perhaps to the $3.5B range when including a launch on the SLS.  Congress will need to substantially increase the planetary budget to prevent the Europa mission from crowding out the smaller planetary missions that provide balance to the program.  While Congress can pass budget laws directing year to year spending, meeting these aggressive goals will require that the President’s Office of Management and Budget (OMB) accepts the new plan and allows NASA to sign the necessary multi-year contracts with its vendors.  In the past, OMB has resisted prioritizing the Planetary Science budget to accommodate a Europa mission.

Second, the driving force behind the expanded mission depends on one Congressman and his continued re-election, his political party’s continued control of Congress, and his health.  The alternative approach would be to run the exploration as NASA has run the Mars program by spreading costs out over a sequence of missions.  This would be in the vein of the proposed “Ocean Worlds” program currently being shopped to NASA and Congress.

I expect that in the next few months that we will learn considerably more about the lander’s design and NASA’s plans on how it will fit into its overall planetary program.

Tuesday, December 22, 2015

InSight's Problems - Possible Impacts

As I'm sure many of you already know, NASA has cancelled the launch of the Mars InSight geophysical mission because of problems with its prime instrument, the seismometer.  Casey Drier with the Planetary Society has a good description of the problem and why NASA's managers decided not to proceed with the planned launch in March.

Artist's conception of the InSight lander with the seismometer and heat flow instruments deployed.  Credit JPL

Deciding not to proceed with the launch just three months away will likely prove to be the toughest or one of the toughest decisions these managers will make in their careers.  I also believe they made the right decision.  Hail Mary passes make for good football but terrible management of space missions.

So how might this affect NASA's planetary program?  The answer breaks into three parts.

In the short term, there are three immediate steps.  The spacecraft will be returned to its manufacturer, Lockheed-Martin for storage.  The French space agency, which is supplying the seismometer, will work to fix the problems to enable the possibility of a future launch.  And the launch vehicle will need to be stored or assigned to launch another spacecraft.

In the intermediate term, NASA's managers must decide whether to launch the mission 26 months later in 2018 when Mars and Earth again align or cancel the mission.  The InSight mission is funded through NASA's Discovery program, which imposes strict cost caps on missions.  The costs of storing the spacecraft and then retesting it prior to a later launch very likely will bust those caps.  NASA's regulations require a formal review of whether or not to delay the mission or cancel it.

For the longer term, InSight's problems could have two effects on the pace of future Discovery missions.  If the mission is cancelled, there are no additional costs associated with it and almost $150M in future InSight costs are avoided.  Five proposed missions are currently being evaluated for selection in approximately nine months.  NASA's managers have stated that they would like to select two of those missions if they possibly can.  If InSight is cancelled, then NASA has additional funds to apply to the next next mission or missions selected next September.

If NASA decides not to cancel InSight, then the space agency will have new costs associated with the storage and later recommissioning of the spacecraft and possibly higher costs for the launch.  In addition, the costs of operating the mission will be pushed forward into 2018 to 2020 when the peak funding for the development of the next Discovery missions is required.  (NASA can't simply bank the money it would have spent on InSight from 2016 to 2018.  That money would either need to be spent on other missions or returned to the general federal budget.  The Federal government works on a spend as you go basis.)  The likely result is that NASA would be able to select just one new Discovery mission this coming September.  One of my correspondents tells me that just last week NASA's manager for its planetary program, Jim Green, said at a scientific conference that he hoped to select two Discovery missions in September unless InSight is delayed.

No matter how you look at it, InSight's problems seem likely to cause NASA to fly one fewer Discovery mission.  Either InSight is cancelled or NASA selects just one Discovery mission from the current competition.

If I were to place a bet, I would guess that NASA will decide to delay InSight, select just one new Discovery mission this September for launch by 2021, and select the next Discovery mission in 2019 for launch in the mid-2020s.  The science InSight would conduct to study the interior of Mars is compelling and the costs of delay seem likely to be less than the cost of a full new Discovery mission.  But please remember, this is only my speculation.  NASA's managers need to make the hard call.

Sunday, November 22, 2015

Justin Atchison has been kind enough to provide background on the proposed NASA Double Asteroid Redirection Test (DART) mission.  As Justin states, the NASA mission may be paired with a European orbiter that would also deliver a lander and several CubeSats to explore both members of this binary asteroid pair.  ESA has a nice webpage describing their proposed mission: Asteroid Impact Mission if you would like additional information.

Preface: Greetings, I’m Justin Atchison, an aerospace engineer at the Johns Hopkins University Applied Physics Laboratory. I’m excited to be guest-writing an article about the Double Asteroid Redirection Test (DART). I recently presented some research about DART at the International Astronautical Congress, which I was able to attend thanks to a travel fellowship through the Future Space Leaders Foundation (FSLF). I’d strongly encourage any student or young-professional (under 35) to apply for this grant next year. It’s a great opportunity to attend this premier conference and interact with a variety of leaders in the aerospace field. FSLF also hosts the Future Space Event on Capitol Hill each summer, which offers engagement with US Congress and aerospace executives on the latest and most relevant space-related topics.


The Asteroid Impact Deflection Assessment (AIDA) mission is a proposed joint program between NASA and ESA to study and demonstrate kinetic asteroid deflection as a means of Earth impact mitigation [[1]]. Basically, we want to understand what’s involved with using a high speed impacting spacecraft to change the orbit of an asteroid that is (hypothetically) threatening Earth.  DART is the proposed NASA component of AIDA. Its role is to hit a small asteroid at high enough velocity to create a measurable deflection. The proposed ESA component of AIDA is called the Asteroid Impact Mission (AIM), and its role is to rendezvous in advance of DART and study the pre- and post-impact asteroid conditions in-situ. The final component of AIDA is Earth-based observations.

AIDA consists of DART, AIM, and Earth Based Observations. AIM will rendezvous with the binary asteroid Didymos and observe as DART impacts the smaller secondary asteroid.

What makes this concept unique is that the target is a binary asteroid system, specifically Didymos (1996 GT). Didymos was one of the first Near Earth binary asteroids to be discovered; we now estimate that roughly 15% of all Near Earth Asteroids (NEAs) are binary systems [[2]]. The primary body of the Didymos system is about 800 m in diameter, and the secondary body (playfully called “Didymoon” on the mission) is only 170 m in diameter [[3]]. We want to impact Didymoon using the DART spacecraft and measure the change in orbit about the primary. 

Here’s the point—the binary system represents a sensitive experimental setup. The DART impactor is about 300 kg and will impact the asteroid at nearly 7 km/s. That is a lot of linear momentum, but we estimate that it will only impart ~1 mm/s to Didymoon. For a single asteroid, the heliocentric velocity is ~30 km/s. Measuring 1 mm/s out of 30 km/s (1 part in 10,000,000) would be exceedingly difficult. You could pick an even smaller target, so that you impart a higher amount of momentum, but then you have the challenge of trying to observe a very small body at long ranges. However, in the case of a binary, the velocity of Didymoon with respect to the primary is only about 10 cm/s. The effective signal-to-background is now just 1 in 1000. (Ok…the order-of-magnitude argument isn’t entirely accurate here, but it communicates the point.) Further, Didymoon is observable at long ranges as a signature in the system’s light-curve. In fact, that’s how it was first confirmed to be a binary system.

Why Didymos?

Of the known binary asteroids, Didymos makes an excellent candidate target for the experiment for a few reasons:

      Accessibility – Didymos falls in a class of asteroids that requires less total V with which to rendezvous than the Moon. That means that AIM can be kept relatively small and affordable. Likewise, there are very feasible trajectories for DART to impact with high relative velocities.

Orbits of Earth and Didymos from above the ecliptic (top), and from the “side” along the ecliptic y-axis [[4]]. The dashed line shows the line where the two orbit planes intersect. Didymos is inclined about 3.4 degrees relative to Earth’s orbit plane.

Conjunction – In October of 2022, Didymos will make a rare (every ~25 years) conjunction with Earth. If we time the encounter to occur during that conjunction, Earth-based instruments, including radar, can participate in the observations and add a great deal of science. 

Earth-Didymos range between 2000 and 2050. The 2022 conjunction is especially close.

The Experiment

Although I described the kinetic impact measurement in terms of velocity, we’re really going to be measuring the change in orbit period of Didymoon. This approach has the added advantage that the measurement sensitivity “integrates” with time. After enough time, the phase of Didymoon will be clearly different from what one would predict without the impact. Within practical limits, the longer we watch, the more clear the effect becomes. (The practical limits are associated with the native acceleration environment and the quality of the pre-impact characterization.) 

At this point, if you’re like me, you’re probably wondering why there’s any uncertainty in the imparted momentum. (In fact, there’s a Reddit thread dedicated to this exact question). After all, didn’t high school physics pretty clearly describe this type of event using inelastic collisions and conservation of linear momentum? The one word answer is: hypervelocity. That’s meant to say that the velocity of the impact, 7 km/s, is significantly higher than the speed-of-sound in an asteroid. Bear in mind that a bullet travels at something like 1 km/s.  When DART impacts the surface, its energy will propagate into the asteroid where it will generate a new crater. Some fraction of the material from the crater will be ejected off the asteroid at high speed. This ejected material acts like a thruster, imparting additional momentum to Didymoon. Total momentum is conserved…but Didymoon gets more than DART’s momentum added to it because it’s ejecting some of its material. 

The amount of momentum amplification is called β and it depends principally on the speed of impact, the asteroid composition, and the asteroid structure (porosity and cohesion) [1]:

(mS and vS are the mass and velocity of the spacecraft. mA is the mass of the bulk asteroid. ∆v is the change in the asteroid’s velocity.)

Estimates suggest that β can be anywhere between 1 and 5 for the types of conditions you’d expect for kinetic deflection. (Yes, it’s hypothetically possible to get cases where β < 1, in the event that material is ejected from the asteroid’s side opposite the impact.)

There are roughly three different cases of expected impact momentum exchange. The low-speed case (at top) matches simple expectations for an inelastic collision (β = 1). The middle hypervelocity case is more likely; a crater is formed at the impact site and debris is ejected away. This ejected material imparts additional momentum to the remaining asteroid, giving an effective “amplification” to the momentum exchange (β > 1). The bottom hypervelocity case is unlikely, but shows a scenario where the impact’s shock wave crosses the asteroid and ejects material on the far side of the asteroid. In this case, some of the spacecraft’s momentum is transferred to the ejected material and the bulk asteroid receives a reduced change in momentum (β < 1).

Right now, the DART team is parametrically modeling the encounter using really impressive numerical simulations. The codes were developed from explosives and weapons modeling, which are some of the most complex computer models to create. These simulations require massive supercomputing resources and help us understand what happens to different kinds of materials as they undergo complex physical events such as high-velocity impacts and explosionsSpecifically, the region around the impact is divided up into a grid consisting of millions of little asteroid “cells”. The simulation evaluates the propagation of energy and momentum through each cell, determining the interaction of each cell with its neighbors. The fraction of cells containing material that are ejected from the surface represents the amount of material that affects the value of β

The simulations are very useful, but it’s quite difficult to validate them in 1-g.  Scientists conduct scaled experiments with high energy mechanical contraptions like air-guns and catapults, but it’s hard to accurately reproduce the expected structural properties of the target--a gravitationally bound “rubble pile”. With that in mind, DART will represent a chance to validate our hypervelocity impact modeling and enable us to then  more-accurately extrapolate to other asteroid types.

Various computational hydrocode results for DART impacts at differing geometries [[5]].

Program Status

Together, the two proposed missions form a coherent experiment. However, they’re being developed such that they aren’t codependent: That is, AIM will be the first spacecraft mission to a binary asteroid, which would be pretty exciting with or without DART.  AIM carries a suite of instruments including a visible imager, thermal infrared imager, two radar systems, and a landed package. It also will be carrying a set of 2-6 cubesats [[6]].

Should only DART launch, the change in orbit period can be observed without AIM by using Earth-based observatories. Amateur and professional astronomers worldwide will want to study the event to try to characterize the post-impact environment. (Think Shoemaker-Levy 9’s impact at Jupiter in 1994.) Between light curves and radar, we intend to measure the orbit period change to better than 10%. If AIM is there, the results are obviously much clearer and more accurate, but nonetheless the experiment can be conducted using DART only.

DART itself is a relatively simple spacecraft with only a single instrument, DRACO, which is an imager derived from the narrow-field-of-view telescope on New Horizons. DART has the challenge of reliably targeting and impacting a slight 170 m diameter target. To successfully achieve this, we’ve been working hard to develop and prove out algorithms for autonomous optical guidance, navigation, and control (GNC). The GNC software must distinguish between the two bodies at Didymos and then drive the spacecraft towards the image centroid corresponding to Didymoon, all within a matter of hours. For this complex problem, we’re leveraging decades of missile guidance algorithms, namely something developed in the 1970’s called Proportional Navigation. I can’t help but call to mind a quote from the opening of the Arthur C. Clarke book Rendezvous with Rama (1973):

“A hundred years earlier, a much poorer world, with far feebler resources had squandered its wealth attempting to destroy weapons launched suicidally by mankind against itself. The effort had never been successful, but the skills acquired then had not been forgotten. Now, they could be used for a far nobler purpose, and on an infinitely vaster stage. No meteorite large enough to cause catastrophe would ever again be allowed to breach the defenses of Earth.”

Lofty quotes aside… AIM is currently a Phase A/B1 study within ESA, with two companies developing conceptual designs. DART is also a Phase A study, managed by the NASA Planetary Defense Coordination Office, within the Science Mission Directorate at NASA Headquarters. Both projects will proceed over the next year towards their respective key decision points. 

I hope you’ll agree that this is a compelling concept. To me, it seems to answer that old criticism, 

“Everyone complains about the weather asteroid-impact-threat, but no one does anything about it.”

Let’s do something about it.


[1] A. F. Cheng, J. A. Atchison, B. Kantsiper, A. S. Rivkin, A. M. Stickle, C. Reed, A. Galvez, I. Carnelli, and P. Michel, “Asteroid Impact and Deflection Assessment Mission," Acta Astronautica, vol. 115, pp. 262-269, 2015.
[2] Bottke W. and H. J. Melosh, Binary Asteroids and the Formation of Doublet Craters, Icarus 124: 372–391 (1996)
[3] Scheirich, P., and P. Pravec, 2009, Modeling of lightcurves of binary asteroids, Icarus, 200:531-547
[4] J. A. Atchison, M. T. Ozimek, B. Kantsiper, and A. F. Cheng, “Trajectory Options for the DART Mission,” International Astronautical Congress, Jerusalem, Israel, IAC-15-C1.1.31080.
[5] A. M. Stickle, J. A. Atchison, O. S. Barnouin, A. F. Cheng, D. A. Crawford, C. M. Ernst, Z. Fletcher, and A. S. Rivkin, “Modeling momentum transfer from kinetic impacts: Implications for redirecting asteroids," 13th Hypervelocity Impact Symposium, 2014.
[6] I. Carnelli, A. Galvez, K. Mellab, M. Kueppers, “Industrial Design of ESA Asteroid Impact Mission,” International Astronautical Congress, IAC-15-A3.4.9.x30901, 2015.