DEFLECTING HAZARDOUS ASTEROIDS FROM COLLISION WITH THE EARTH BY USING SMALL ASTEROIDS
IAAPDC130416P
David Dunham^{(1,3)}, Natan Eismont^{(2,3)}, Michael Boyarsky^{(2,3)}, Anton Ledkov^{(2,3)}, Ravil Nazirov^{(2,3)}, Eugene Chumachenko^{(3,2)} and Konstantin Fedyaev^{(2,3)}
^{(1)}KinetX, Inc., 7913 Kara Ct, Greenbelt, MD 20770, USA, +13015265590, david.dunham@kinetx.com
^{(2)}Space Research Institute of Russian Academy of Science, 84/32 Profsoyuznaya Str., 117997, Moscow, Russia, +7 916 628 6139, NEismont@iki.rssi.ru
^{ (3}^{)} National research university ‘Higher school of economics’, per. Trekhsvyatitelskiy B., d. 3,
109028, Moscow, Russia, +74959170750, kommek@miem.edu.ru
Abstract: The main idea of the proposed approach consists of targeting a very small asteroid to impact a larger dangerous one. The minimum size of this small asteroid is determined by the ability to detect it and to determine its orbit. The small object may have a diameter of about 10 15 meters. Asteroids are selected from the nearEarth class that have a flyby distance from Earth of the order of hundreds of thousands of kilometers. According to current estimates, the number of near Earth asteroids with such sizes is high enough. So there is a possibility to find the required small asteroid. Further, the possibility is evaluated of changing the small asteroid’s orbit so that by application of a very limited deltaV impulse to the asteroid, the latter is transferred to a gravity assist maneuver (Earth swingby) that puts it on a collision course with a dangerous asteroid. It is obvious that in order to apply the required V pulse it is necessary to install on the small asteroid an appropriate propulsion system with required propellant mass. A control system similar to that used on a spacecraft is also necessary. The Keck study [1] has already shown the feasibility of performing the above steps to change the orbit of similar asteroids to target them to a lunar orbit; our idea is similar in orbital energetics to that, only for slightly larger objects and targeting them to an Earth Bplane point rather than the vicinity of the Moon. Of course, any real test of this (or of any other deflection) strategy should first be performed on a benign asteroid whose orbital parameters give the asteroid no significant chance of a natural impact with the Earth during at least the next million years, even in the case of small changes in its orbit.
The main goal of the paper is to demonstrate that this concept is feasible. For this purpose, studies were performed to optimize the whole list of possible missions, considering all known NEA’s regardless of their mass that might be used to deflect a hazardous asteroid from collision with the Earth. An asteroid that is too large can be considered since just a boulder of the right size could be lifted from its surface and used. The criterion of optimization is the total mass of the spacecraft to be launched from low Earth orbit in order to deflect the dangerous asteroid from hitting the Earth. For this the candidate small asteroids were selected whose targeting to the near Earth gravity assist maneuver needs a very limited V impulse and where the total V to place the spacecraft with the required amount of propellant on the surface of the asteroid is minimal. For our study, we have selected Apophis [2] as the target hazardous asteroid just as an example, since that asteroid is not now actually expected to pose a significant threat in the next decades. All operations are supposed to be completed before the end of 2035. During the studies aimed to choose the appropriate asteroid to be directed to the dangerous near Earth object, it was found that in the available catalogue, there are candidates satisfying our needs, including some that need a V pulse of less than 10 m/s to strike the hazardous object. One asteroid was found needing only 2.5 m/s in order to transfer it to the necessary Earth swingby Bplane point to impact Apophis. Previous studies showed that using a small asteroid as a projectile with an Earth gravity assist, as described more below, is about two orders of magnitude more efficient than a direct kinetic impact using a spacecraft [3].
Optimal transfer trajectories to Apophis were found for the asteroids in the list of candidates to be targeted to Apophis. The V required to land a spacecraft on the asteroid’s surface starting from an initial low Earth orbit was used as an optimization criterion. It was shown that its value does not exceed 6 km/s.
The problem of trajectory correction maneuvers is explored in the paper; they are strongly dependent on the achievable accuracy of the orbit parameters determined from processing the tracking data. In order to reach the best possible accuracy in targeting a small asteroid to Apophis, a beacon spacecraft is used on the surface of Apophis.
Different types of engine units that can deliver the required nominal and trajectory correction maneuvers of the small asteroid are compared, including standard chemical and electric propulsion engines, and also mass drivers that can accelerate material from the asteroid to generate thrust.
Keywords: NearEarth asteroids, planetary protection
1. Introduction
We want to answer the question: Do small asteroids exist in the contemporary catalogue that can be used to transfer onto a trajectory of collision with Apophis before that asteroid approaches the Earth in April 2036, assuming that we can use only launch vehicles and technologies of spacecraft motion control that are available now. We mean realizing a scenario such that a spacecraft with propellant of enough mass is sent to chosen small asteroidprojectile, then lands on that asteroid and, after its attachment to the surface, the spacecraft engine gives to the asteroid a velocity pulse, which transfers the asteroid to the trajectory of its collision with Apophis using an Earth gravity assist maneuver. Results are given below to give some preliminary confirmation that the described method of deflecting Apophis from a trajectory of possible collision with the Earth is feasible. Taking into account the latest successes in discovering new near Earth asteroids, one may state that the the chances to find an asteroidprojectile needed to implement the proposed method of Earth defense will be improving.
2. Gravity assist maneuver as a tool to target asteroidprojectile to hazardous near Earth object.
Asteroids which are considered as small have mass about 1500 tons so to control their motion in classical use of this word i.e. by applying pulse of rocket engine in order to change their velocity by several kilometers per second is hardly possible. But if a gravity assist maneuver is used as the tool to amplify changes of orbital parameters with a small velocity pulse (deltaV) allows changing the pericenter height of the controlled body (asteroidprojectile) when it flys by the planet (Earth) by a value sufficient to turn the relative velocity vector by dozens of degrees, then we have an extremely efficient tool of orbital control for the celestial body having so large a mass as compared with the usual spacecraft.
By selection of a relative velocity position at infinity with respect to the flyby planet (with fixed pericenter radius) we receive any required plane of relative orbit with appropriate direction of relative velocity vector after the flyby. This is illustrated in Fig.1 where a cylinder of possible vectors of relative velocities (at infinity) at arrival, and the resulting postflyby cone of velocity vectors of departure are shown.
Figure 1. a) Cylinder of possible vectors of relative velocities at arrival, and resulting postflyby cone of velocity vectors of departure, b) Geometry of gravity assist maneuver in coordinate system connected with the Sun
The geometry of a gravity assist maneuver is illustrated in Figure 1b. In Fig. 1b V_{p }is the velocity vector of the planet with a flyby during a gravity assist maneuver (the Earth in our case) given in coordinate system connected with the Sun, V_{a }is arrival velocity vector of asteroid in the same reference (coordinate) system and V_{0 } is the velocity vector of the asteroid relative to the planet at the moment of its arrival at the Earth referred to infinity (if gravity of the Earth would be equal to zero this vector would be equal to the one we consider). After planet flyby this vector is turned by an angle α (the bend angle) which can be calculated using the following formula [4]:
(1)
where:
r – pericenter radius, V – relative velocity of asteroidprojectile at infinity, μ – gravitational constant of the planet.
As one can see from the formula, with decreasing the pericenter radius, the α angle of relative velocity vector turn due to the flyby increases, reaching 180 degrees if the pericenter radius reaches zero. Thus if there are no limits constraining the lowest value of the pericenter radius, then the asteroid’s relative velocity vector can be reversed, to opposite its initial direction. The whole family of the departure relative velocity vectors forms in this case the sphere with V radius. The asteroid’s velocity vector in the heliocentric reference system is the sum of the planet’s velocity vector and the relative vector of the asteroid with respect to the planet (Earth) . Thus the asteroid’s velocity vector in the heliocentric coordinate system may be any vector with initial point in the same point as the planet’s vector and any final point on the mentioned sphere. But this is valid only in case if the planet’s radius is zero. If one will take into account the radius of the planet, then only part of the surface of the sphere can be reached after the gravity assist maneuver. This part is constrained by the cone with the axis coinciding with the arriving vector of asteroid relative velocity V_{0}. The semiangle of this cone is α_{max, }calculated by use of the formula given above where the pericenter radius is equal to the minimum allowed one. With increasing possible radius of pericenter of the flyby trajectory from zero to infinity, the area of achievable departure velocity vectors shrinks from the whole sphere to the one point which is the end point of the arriving velocity vector. Accordingly, the area of achievable velocity vectors of planet flyby body (asteroidprojectile) is decreased.
3. Lambert problem as the principal tool to design trajectories design using gravity assist maneuvers.
A scenario of a mission to deflect a hazardous near Earth object from an Earth collision trajectory consists of the following phases illustrated in Figure 2:
Figure 2. Trajectories of Earth, Apophis, asteroidprojectile 2011 UK and spacecraft transfer trajectory from Earth to 2011 UK
It is obvious that during the whole mission all available tracking instruments and facilities are needed to solve the navigation tasks with maximum achievable accuracy. The importance of this part of mission operations is much higher than in case of the usual spacecraft motion control because correction maneuvers must be kept small, due to the large mass of the projectileasteroid and the limited amount of fuel available to change its orbital elements. The task of the mission design at large consists in the choice of all available free parameters in such a way that, with the maximum payload possible using currently available launch vehicles, results in the maximum deviation of the velocity vector of the target asteroid (Apophis) after it is hit by the controllable asteroidprojectile. This task is multi parametrical so it is to be solved in several phases.
The first step is to choose an optimal transfer of the spacecraft from low Earth orbit (LEO) to the asteroid chosen as the asteroidprojectile. For this the Lambert problem is to be solved with the goal to minimize the required deltaV to execute this mission. As it is well known, the Lambert problem [5, 6] consists in the choice of initial orbital parameters that during a given time, allows transfer of a zeromass object along a Keplerian orbit from one specified point to another. Our goal is to choose from possible dates of launch and arrival, the optimal ones in terms of total deltaV. During this phase acceptable candidate asteroids are to be chosen as the dates of their departure and arrival. It should be mentioned that on this phase of calculations of asteroid motion the gravity field is centered at the Sun. The coordinates of Earth and coordinates of asteroids are taken from appropriate catalogues using the SPICE system [7].
The next step is to find a transfer orbit from the Earth to Apophis with date and time of start the same as the ones of arrival to the Earth of the asteroidprojectile. The magnitude of the velocity vector relative to Earth is chosen to be the same as the magnitude of the arriving velocity vector. As a result of this second (modified) Lambert problem solution, we receive the departure relative velocity vector and transfer trajectory for the asteroidprojectile at large. Practically it was confirmed that the standard algorithm of Lambert problem solution may be used for both parts of the trajectory: for the part approaching the Earth and for the part from the Earth to Apophis. For this it is enough to minimize the sum of the deltaV for transfer of the asteroidprojectile from its initial orbit to the orbit reaching the Earth, and the difference of magnitudes of the arriving and departing relative velocities at the Earth. As calculations showed the result of such minimization is zero difference of these magnitudes that exactly correspond to the demand of mutual matching of these trajectories (that is, an unpowered Earth swingby), so we receive one uninterrupted trajectory of asteroidprojectile from the first correction velocity pulse to the collision with asteroid Apophis.
With the use of the described method and the catalogue of the solar system bodies orbital parameters contained in SPICE calculations, we have fulfilled the aim to select the candidate asteroids for targeting them to Apophis with the use of a gravity assist maneuver near Earth. The criterion of selection of these asteroids was the required deltaV to transfer the asteroid from its original orbit to the trajectory of collision with Apophis after the gravity assist maneuver near Earth. The sizes of the asteroids were also taken into account as the necessary deltaV to deliver the spacecraft onto the surface of the asteroidprojectile. The best five asteroids, satisfying the described requirements that were chosen by our studies, and their key characteristics, are given in Table 1.
Table 1. Results of the selected candidate asteroids and the orbit designs
Asteroid

2006 XV4

2006 SU49

1997 XF11

2011 UK10

1994 GV

DeltaV value, m/s

2.38

7.89

10.05

15.94

17.72

Perigee radius, km

16473.19

15873.40

42851.84

31912.94

7427.54

Velocity in perigee with respect to Earth, km/s

9.61

5.03

14.08

8.98

13.37

Velocity vector bend angle of relative to the Earth, deg.

23.98

59.78

5.14

21.14

50.85

Date of maneuver execution

2029/03/17

2027/06/11

2027/04/27

2025/09/13

2028/09/12

Date of perigee reaching

2031/12/11

2029/01/23

2028/10/26

2026/10/10

2031/04/13

Date of collision of asteroidprojectile with Apophis

2034/04/08

2029/10/06

2030/08/06

2027/08/06

2031/12/24

Impact velocity with Apophis, km/s

15.3

4.9

11.0

2.3

14.1

Magnitude

24.87

19.54

16.9

24.91

27.46

Size of asteroidprojectile

25 ≈ 60 m

330 ≈ 750 m

1 ≈ 2 km

25 ≈ 60 m

8 ≈ 19 m

V^{2} at infinity after s/c launch from LEO, km^{2}/s^{2}

63.128
(3.7*)

0.36

6.447

47.182
(1.488*)

30.128
(2.427*)

DeltaV of braking for landing S/C on an asteroid, km/s

0.89
(9.6*)

4.67

7.89

0.543
(5.571^{*})

0.543
(6.860*)

∆V_{t }, km/s (total)

6.69





5.744

5.954

*for departure deltaV optimization
4. Optimization of spacecraft transfer trajectory with start from low Earth orbit (LEO) and landing on asteroidprojectile surface.
A natural criterion of optimization for choice of the trajectory of spacecraft delivery to the asteroidprojectile surface is the maximum mass of the spacecraft after landing. For our studies it is enough to use instead of this criterion, one very close to it: The total ∆V_{t} required to start from LEO and to execute the maneuver to land the spacecraft on the asteroid’s surface. For calculating first constituent (∆V_{s}) of the deltaV we assume that the spacecraft starts from the circular LEO with 200 km height. The second constituent we estimate as the relative velocity of spacecraft with respect to asteroid when arriving at it (rendezvous deltaV).
In order to minimize available software [8, 9, 10] modification, another approach for solving the problem was used: we minimized the function F = W_{1}*C_{3}+W_{2}*∆V_{a}, where C_{3} is the square of relative asymptotic velocity with respect to Earth of the trajectory departing to the asteroid, ∆V_{a} is the asteroid rendezvous deltaV, and W_{1} and W_{2} are the weight factors for the procedure used to minimize the function. Several values of W_{2 }were tested keeping W_{1 }=1 for all cases. It is obvious that with increasing the W_{2} value the influence of arriving velocity on the solution is raised. In the presented table of received solutions that includes the data of start, data of arrival, square of asymptotic velocity at departure, deltaV required for departure, and asteroid rendezvous deltaV. Table 2 presents key orbital parameters for the transfer mission of the spacecraft to be landed on 2011 UK10 asteroid calculated by the method of optimization for different weight factors W_{1}, W_{2}. The search of optimal trajectory was done for the spacecraft departure inside an interval of dates beginning with departure not earlier than 20200101 and ending by arrival not later than 20250815.
Table 2. Key orbital parameters for transfer mission of the spacecraft to asteroidprojectile

W_{2}

Optimal time of departure from Earth

Optimal time of arrival at 2011 UK10

C_{3},
km^{2}/s^{2}

∆V_{a} ,
km/s

Duration of transfer, days

∆V_{s },
km/s

∆V_{t },
km/s

2011 UK10

20

2022/10/13

2023/12/09

47.1824

0.54275

422.5234

5.201

5.744

1994 GV

20

2025/12/03

2028/01/02

50.6314

0.5913

1056.6681

5.333

5.924

2006 XV4

20

11/21/2023

03/28/2027

63.1283

0.89119

1222.5468

5.8

6.69

As one can see from the table the last column presents the optimal trajectory with a total deltaV
of 5.744 km/s. The impulse required to rendezvous with the asteroid is only 0.5427 m/s. This trajectory is shown in Fig.3, an ecliptic projection with the orbits of inner planets and orbit of 2011 UK asteroid. Rough estimations show that a standard Proton launch vehicle with Breeze upper stage using optimal transfer trajectory can deliver to the surface of 2011 UK10 asteroid (after consuming a total deltaV of 5.744 km/s) a payload with 2300 kg mass.
Figure 3. Optimal transfer trajectory to 2011 UK asteroid with minimum total deltaV
5. Comparison of Lambert problem solution and results of numerical integration of differential equations of Solar System bodies
For 2011 UK10 (#3582088 in SPICE system), a comparison was made of results of numerical integration of the equations of motion, taking into account solar system gravity field generated by the Sun and planets on the asteroid, with the ones calculated by Lambert problem solution.
Table 3. trajectory parameters corresponding minimum deltaV

Use Lambert problem

Numerically integrated trajectory

Numerically integrated trajectory with optimization by choice of T_{1 }and_{ }T_{3}

Date and time of maneuver execution  T_{1}

20250913 10:37

20250913 10:37

20250824 10:37

DeltaV value, m/s

15.9

22.0

15.6

Date and time of perigee passage  T_{2}

20261010 13:07

20261010 07:06

20261010 13:02

Perigee radius, km

31912.7

31578.7

31683.7

Velocity in perigee with respect to Earth, km/s

8.98

8.977

8.995

Date and time of collision of asteroidprojectile with Apophis  T_{3}

20270806 07:13

20270806 08:54

20270807 10:27

Apophis impact velocity, km/s

2.25

2.26

2.28

Thus the data given in the above list of key trajectory parameters confirms that we can use conic Lambert problem solutions for accurateenough calculations for calculating the possibilities of the use of small asteroids to deflect dangerous ones from their original trajectory that might collide with the Earth.
6. Analysis of results for ideal cases
The figures given in Table 1 and presented by key parameters of trajectories found by our studies are for ideal cases, i.e. for the ones when the trajectory is nominal with no (zero) deviation of the real trajectory from the calculated trajectory. But for the considered missions the part of propellant to be consumed during the mission for correction maneuvers may be comparable with those for the nominal maneuvers.
If one considers the mission effectiveness only in terms of the nominal trajectory, then from the 5 cases presented in Table 1, the most promising case is the mission using the asteroid 1994 GV. The main argument for such a choice is the estimated mass of the asteroids in Table 1. The asteroids closest in size are 2011 UK10 and 2006 XV4 are supposed (if their densities are the same) to have a mass by a factor 30 times higher. Assuming that mass of 1994 GV asteroid is 1350 tons (radius 6 meters, density 1.5 t/m^{3}), in order to change its velocity by the needed 17.72 m/s using engine unit with specific impulse 3300 m/s, one needs to consume 7.23 tons of propellant. It means that 4 Proton launch vehicles are to be used in order to deliver the required amount of propellant to the asteroid surface. It is possible but not so simple. In addition some consumption of propellant is necessary for trajectory correction maneuvers. If one refers to the Deep Impact mission to estimate the deltaV needed for these maneuvers, then the expected figures may reach a few dozen meters per second [11]. It means that for implementation of the proposed technology of dangerous asteroid deflection we need to find more asteroids satisfying our demands in terms of their size and required deltaV to target them to dangerous objects like Apophis. For example if 2006 XV4 would have the same size as 1994 GV and similar required deltaV to reach it then the required mass of propellant to target it to Apophis would be only 0.98 tons if one would not take into account the propellant needed for correction maneuvers. That might be done in any case by using a suitablysized boulder lifted from the surface of 2006 XV4, rather than trying to move the whole asteroid.
We note that even with the use of 10 Proton launches to accomplish the proposed plan for this selected case, it is 27 times more effective than a direct kinetic impact of the spacecraft to Apophis. Correction maneuvers might be decreased significantly by the use of a transponder to be delivered to the surface of Apophis before execution of the deflection mission.
7. Resonance orbits as the basis for constructing a planetary defense system.
The proposed concept of using small asteroids to deflect hazardous objects from the trajectory of collision with Earth may be developed further. The idea is to transfer small asteroids onto Earth resonance orbits, for example with period of one year, using the Earth gravity assist method described above. Our preliminary studies have confirmed that it is possible to find 11 asteroids which are possible to transfer to such orbits with a deltaV not exceeding 20 m/s. Thus a system is constructed which allows sending asteroidprojectile to the hazardous object approximately each month during year. This is like a combination of two previous ideas, the use of nuclear “Soldier” spacecraft in Earthreturn orbits as part of the SHIELD concept [12], and stationing small asteroids in quasiperiodic orbits about the collinear SunEarth Lagrange points for use as “David’s stones” against threatening “Goliath” asteroids [13].
8. Conclusions
The described method of deflecting dangerous asteroids from a collision trajectory with the Earth seems to be feasible, as shown using Apophis as an example. It was found that only a very small deltaV (2.38 m/s) may be needed to transfer a small asteroid to a trajectory with an Earth gravity assist maneuver, followed by collision of this asteroid with a hazardous object like Apophis. The proposed method allows a velocity change of a dangerous object by values greater than by any other contemporary technology. For practical implementation of the proposed approach, some further progress in broadening the catalogue of candidate asteroidprojectile is needed, especially for small asteroids. Also additional studies are needed to decrease correction maneuver deltaVs.
9. References
[1] Brophy, J., Culick, F., Dimotakis, P., and Friedman, L., “A Safe Stepping Stone Into the Solar System.” Paper IAC12.A5.4.11, presented at the 63^{rd} International Astronautical Congress, Naples, Italy, October 2012.
[2] JPL SmallBody Browser: 99942 Apophis (2004 MN4), http:/ssd.jpl.nasa.gov/sbdb.cgi?sstr=99942
[3] Nazirov R., Eismont N. “Gravitational Maneuvers as a Way to Direct Small Asteroids to Trajectory of a Rendezvous with Dangerous NearEarth Objects”. Cosmic Research, 2010, Vol.48, No.5, pp. 479484
[4] Spacecraft Attitude determination and Control, Wertz, J.R., Ed., D.Reidel Publishing Company, 1985., p. 60.
[5] Lancaster E.R., Blanchard R.C., “A unified form of Lambert's theorem.” NASA technical note TN D5368,1969.
[6] Gooding R.H., “A procedure for the solution of Lambert's orbital boundaryvalue problem.” Celestial Mechanics and Dynamical Astronomy (ISSN 09232958), 48, № 2:145165, 1990.
[7] “NAIF/SPICE ancillary information system”, http://naif.jpl.nasa.gov/naif/about.html
[8] Rody P.S. Oldenhuis, “Robust solver for Lambert's orbitalboundary value problem”, http://www.mathworks.com/matlabcentral/fileexchange/26348robustsolverforlambertsorbitalboundaryvalueproblem
[9] Izzo D., ESA Advanced Concepts team, http://www.esa.int/gsp/ACT/inf/op/globopt.htm
[10] Adam Harden, “Trajectory Optimization Tool”, http://www.orbithangar.com/searchid.php?ID=5418
[11] Lakdavala, E, “Deep Impact Successfully Splits in Final Hours before Comet Encounter, July 3, 2005”, Planetary News: Asteroid and Comets, 2005. http://www.planetary.org/news/2005/0703_DeepImpact_Successfully_Splits_in.html
[12] Gold, R. E., “SHIELD: A Comprehensive EarthProtection Architecture”, Adv. Space Res. Vol. 28, No. 8, pp. 11491158, 2001.
[13] Massonnet, D. and Meyssignac, B., “A captured asteroid: Our David’s stone for shielding earth and providing the cheapest extraterrestrial material”, Acta Astronautica Vol. 59, pp. 7783, 2006. 