Far-Field Propagatin of Airburst Events Using a … Propagation of Airburst Events Using a Cartesian...

PLPL

ANET

ARY DEFENSE INTEGRATED

PRO

DU

CT TEAM

Michael J. AftosmisApplied Modeling and Simulation BranchNASA Advanced Supercomputing DivisionNASA Ames Research Center

7-9 July, 2015

Far-Field Propagation of Airburst Events Using a Cartesian Method

1

1st International Workshop on Potentially Hazardous Asteroids Characterization, Atmospheric Entry and Risk Assessment

Acknowledgements

• NASA Advanced Supercomputing Division – Task 3 & 4 teams" Marian Nemec Donovan Mathias

Jonathan Chiew George Anderson Chris Mattenberger Darrel Robertson

Lorien Wheeler

• Entry Systems Division – Task 2 team" " Dinesh Prabhu Ethiraj Venkatapathy

• New York University" " Marsha Berger

2

ARC Planetary Defense IPT

• “Task 3” of the PD IPT

• Focus on ground effects modeling• Airburst & atmospheric propagation

• Surface overpressure & wind prediction

• Ground damage

• Tsunami propagation

“Task 3”

• Inputs come from entry and airburst modeling in Task 2

• Outputs of atmospheric propagation feed tsunami modeling

• Outputs of atmospheric & tsunami modeling feed physics-based risk models in Task 4

3

Ground & Water Effects

ARC Planetary Defense IPT Cart3D

510152025303540

Altit

ude

(km)

5

10

15

20

25

30

35

40

Altit

ude

(km)

510152025303540

Altit

ude

(km)

Mach contours

Mach contours

Elapsed time = 5 sec

time = 35 sec

time = 95 sec

Figure 2. Simulation results for Chelyabinsk meteor using deposition profile from Brown et al.3 18� entryangle, with peak brightness at 29.5 km altitude. Total Energy = 520 kt.

4 of 9

American Institute of Aeronautics and Astronautics

4

• Far-field atmospheric propagation drives

• Ground footprint and land damage prediction

• Atmospheric forcing for tsunami modeling

• Focus

• Perform detailed reconstruction of specific events

• Perform parametric studies to develop surface footprint models for PRA

• Goal is to do thousands of such simulations – need to control computational expense

Goal of atmospheric propagation is prediction of surface footprint

Current work focuses on airburst only, no ground impact

Overview

• Modeling tools & solver

• Verification & Validation• Basic• Chelyabinsk Case Study

• Investigations of ground-footprint sensitivity• Line-source vs time-dependent entry• Entry Angle

• Upcoming Efforts

Report current status of effort and connection with PRA and tsunami

5

Overview






6

• Atmosphere model

• Governing equations

• Solver and simulation methodology

• Model for deposition of mass, momentum & energy

ModelingInviscid scale-height atmosphere model

Geo

met

ric A

ltitu

de (k

m)

Temperature (K)

Pressure (kN/m2)

0

10

20

30

40

50

60

70

80

900 50 100 150

300K2001000

Temperature (K)

Geo

met

ric A

ltitu

de (k

m)

Temperature (K)

Pressure (kN/m2)

0

10

20

30

40

50

60

70

80

900 50 100 150

300K2001000

Temperature (K)

ISO 2533:1975 Isothermal• Atmosphere model based on 1976 Standard Atmosphere (ISO 2533:1975)

• Isothermal approximation for scale-height description

• Use H = 8, and initialize simulations with atmosphere in hydrostatic equilibrium

P (z) = P�e�z/H

⇢(z) =P (z)

RT

7


Geo

met

ric A

ltitu

de (k

m)

Temperature (K)

Pressure (kN/m2)

0

10

20

30

40

50

60

70

80

900 50 100 150

300K2001000

Temperature (K)

Geo

met

ric A

ltitu

de (k

m)

Temperature (K)

Pressure (kN/m2)

0

10

20

30

40

50

60

70

80

900 50 100 150

300K2001000

Temperature (K)




P (z) = P�e�z/H

⇢(z) =P (z)

RT

8


Geo

met

ric A

ltitu

de (k

m)

Temperature (K)

Pressure (kN/m2)

0

10

20

30

40

50

60

70

80

900 50 100 150

300K2001000

Temperature (K)

Geo

met

ric A

ltitu

de (k

m)

Temperature (K)

Pressure (kN/m2)

0

10

20

30

40

50

60

70

80

900 50 100 150

300K2001000

Temperature (K)




P (z) = P�e�z/H

⇢(z) =P (z)

RT

8

Density Contours Temperature Contourst = 0

t = 100

t = 200

t = 300

t = 0

t = 100

t = 200

t = 300

Simple Buoyancy Test

TemperatureDensity

• Use 3D Euler eqs. in strong conservation law form, including body force due to gravity

ModelingInviscid scale-height atmosphere model in hydrostatic equilibrium

U = (⇢, ⇢u, ⇢v, ⇢w, ⇢E)T

S =

0

BBBB@

000

�⇢g�⇢wg

1

CCCCA

d

dt

Z

⌦U dV +

I

@⌦(F · n) dS =

Z

⌦S dV

F =

0

BBBB@

⇢u ⇢v ⇢w⇢u2 + p ⇢uv ⇢uw⇢uv ⇢v2 + p ⇢vw⇢uw ⇢vw ⇢w2 + p

u(⇢E + p) v(⇢E + p) w(⇢E + p)

1

CCCCA

• The state vector of conserved variables is

• Flux density tensor and gravitational body force term are

9

• Original development 1998-2002

• Fully-automated mesh generation for complex geometry

• Unstructured Cartesian cells

• Fully-conservative finite-volume method

• Multigrid accelerated 2nd-order upwind scheme

• Excellent scalability through domain decomposition

• Broad use throughout NASA, US Government and industry• Over 500 users in aerospace community• One of NASAs most heavily used

production solvers, large validation database

Solver: Cart3D OverviewProduction solver based on cut-cell Cartesian mesh method

10

Cart3D

• Original development 1998-2002

• Fully-automated mesh generation from watertight geometry

• Unstructured Cartesian cells

• Fully-conservative finite-volume method

• Multigrid accelerated 2nd-order upwind scheme

• Excellent scalability through domain decomposition

• Broad use throughout NASA, US Government and industry• Over 500 users in aerospace community• One of NASAs most heavily used

production solvers, large validation database

Solver: Cart3D OverviewProduction solver based on cut-cell Cartesian mesh method

16 32 64 128 256 512 1024 2048Number of Cores

4

8

16

32

64

128

256

512

1024

Runt

ime

(sec

onds

)

Distributed Memory (MPI) NAS Pleiades systemShared memory (baseline): NAS Endeavour systemIdeal

2014/01/3057M cell mesh, LM-1021

11

Deposition of Mass, Momentum & Energy

• Focus on ground footprint, not near-field physics

• Abstract entry physics as simply sources of mass, momentum & energy

• Drive simulations via deposition profile taken from:• Models (e.g. ReVelle, Ceplecha, H&G, Shuvalov)• Simulations (Task 2, CTH, ALE3D, Shuvalov, Boslough)• Light-curve derived profiles (Jenniskins, Popova)• Infrasound based energy deposition (Brown, ReVelle)

12

Goal is accurate prediction of surface effects from energy deposition inputs

• Need to derive source terms from deposition profiles

• Release energy, mass and momentum into a corridor of known radius, r

• Over each time step, ∆t, themeteor travels a distance d

• Given: energy deposition profile as a function of altitude

• From modeling

• From simulation

• From observational data

Deposition of Mass, Momentum & EnergyDerive source terms through conservation analysis

dr

13

LETTERdoi:10.1038/nature12741

A 500-kiloton airburst over Chelyabinsk and anenhanced hazard from small impactorsP. G. Brown1,2, J. D. Assink3, L. Astiz4, R. Blaauw5, M. B. Boslough6, J. Borovicka7, N. Brachet3, D. Brown8, M. Campbell-Brown1,L. Ceranna9, W. Cooke10, C. de Groot-Hedlin4, D. P. Drob11, W. Edwards12, L. G. Evers13,14, M. Garces15, J. Gill1, M. Hedlin4,A. Kingery16, G. Laske4, A. Le Pichon3, P. Mialle8, D. E. Moser5, A. Saffer10, E. Silber1, P. Smets13,14, R. E. Spalding6, P. Spurny7,E. Tagliaferri17, D. Uren1, R. J. Weryk1, R. Whitaker18 & Z. Krzeminski1

Most large (over a kilometre in diameter) near-Earth asteroids arenow known, but recognition that airbursts (or fireballs result-ing from nuclear-weapon-sized detonations of meteoroids in theatmosphere) have the potential to do greater damage1 than prev-iously thought has shifted an increasing portion of the residualimpact risk (the risk of impact from an unknown object) to smallerobjects2. Above the threshold size of impactor at which the atmo-sphere absorbs sufficient energy to prevent a ground impact, most ofthe damage is thought to be caused by the airburst shock wave3, butowing to lack of observations this is uncertain4,5. Here we report ananalysis of the damage from the airburst of an asteroid about19 metres (17 to 20 metres) in diameter southeast of Chelyabinsk,Russia, on 15 February 2013, estimated to have an energy equivalentof approximately 500 (6100) kilotons of trinitrotoluene (TNT,where 1 kiloton of TNT 54.18531012 joules). We show that a widely

referenced technique4–6 of estimating airburst damage does notreproduce the observations, and that the mathematical relations7

based on the effects of nuclear weapons—almost always used withthis technique—overestimate blast damage. This suggests that earl-ier damage estimates5,6 near the threshold impactor size are toohigh. We performed a global survey of airbursts of a kiloton or more(including Chelyabinsk), and find that the number of impactorswith diameters of tens of metres may be an order of magnitudehigher than estimates based on other techniques8,9. This suggestsa non-equilibrium (if the population were in a long-term collisionalsteady state the size-frequency distribution would either follow asingle power law or there must be a size-dependent bias in othersurveys) in the near-Earth asteroid population for objects 10 to50 metres in diameter, and shifts more of the residual impact riskto these sizes.

1Department of Physics and Astronomy, University of Western Ontario, London, Ontario N6A 3K7, Canada. 2Centre for Planetary Science and Exploration, University of Western Ontario, London, OntarioN6A 5B7, Canada. 3Commissariat a l’Energie Atomique,DepartementAnalyseSurveillanceEnvironnement (CEA/DAM/DIF), Bruyeres-le-Chatel, 91297 ArpajonCedex, France. 4Laboratory for AtmosphericAcoustics, Institute of Geophysics and Planetary Physics, University of California, San Diego, La Jolla, California 92093-0225, USA. 5Marshall Information Technology Services (MITS)/Dynetics TechnicalServices, NASA Marshall Space Flight Center, Huntsville, Alabama 35812, USA. 6Sandia National Laboratories, PO Box 5800, Albuquerque, New Mexico 87185, USA. 7Astronomical Institute, Academy ofSciences of the Czech Republic, CZ 251 65 Ondrejov, Czech Republic. 8International Data Center, Provisional Technical Secretariat, Comprehensive Test Ban Treaty Organization, PO Box 1200, A-1400Vienna, Austria. 9Bundesanstalt fur Geowissenschaften und Rohstoffe, Stilleweg 2, 30655 Hannover, Germany. 10Meteoroid Environments Office, EV44, Space Environment Team, Marshall Space FlightCenter, Huntsville, Alabama 35812, USA. 11Space Science Division, Naval Research Laboratory, 4555 Overlook Avenue, Washington DC 20375, USA. 12Natural Resources Canada, Canadian HazardInformation Service, 7 Observatory Crescent, Ottawa, Ontario K1A 0Y3, Canada. 13Seismology Division, Royal Netherlands Meteorological Institute, Wilhelminalaan 10, 3732 GK De Bilt, The Netherlands.14Department of Geoscience and Engineering, Faculty of Civil Engineering and Geosciences, Delft University of Technology, Stevinweg 1, 2628 CN Delft, The Netherlands. 15Infrasound Laboratory,University of Hawaii, Manoa 73-4460 Queen Kaahumanu Highway, 119 Kailua-Kona, Hawaii 96740-2638, USA. 16ERC Incorporated/Jacobs ESSSA Group, NASA Marshall Space Flight Center, Huntsville,Alabama 35812, USA. 17ET Space Systems, 5990 Worth Way, Camarillo, California 93012, USA. 18Los Alamos National Laboratory, EES-17 MS F665, PO Box 1663 Los Alamos, New Mexico 87545, USA.

Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b

Figure 1 | Light curve of the Chelyabinsk airburst. a, The brightness profilefor the Chelyabinsk airburst, based on indirect illumination measured fromvideo records. The brightness is an average derived from indirect scattered skybrightness from six videos proximal to the airburst, corrected for the sensorgamma setting, autogain, range and airmass extinction, following theprocedure used for other airburst light curves generated from video24,25. Thelight curve has been normalized using the US government sensor data peakbrightness value of 2.7 3 1013 W sr21, corresponding to an absoluteastronomical magnitude of 228 in the silicon bandpass. The individual videolight curves deviate by less than one magnitude between times 22 and 11.5with larger deviations outside this interval. Time zero corresponds to03:20:32.2 UTC on 15 February 2013. b, The energy deposition per unit height

for the Chelyabinsk airburst, based on video data. The conversion to absoluteenergy deposition per unit path length assumes a blackbody emission of6,000 K and bolometric efficiency of 17%, the same as the assumptions usedto convert earlier US government sensor information to energy26. The heightsare computed using the calibrated trajectory10 and features of the lightcurves common to different video sites, resulting in a height accuracy ofabout 1 km. The total energy of the airburst found by integrating under thecurve exceeds 470 kt. The half-energy-deposition height range is 33–27 km;these are the heights at which energy deposition falls below half thepeak value of approximately 80 kt per kilometre of height, which is reachedat an altitude near 29.5 km.

2 3 8 | N A T U R E | V O L 5 0 3 | 1 4 N O V E M B E R 2 0 1 3

Macmillan Publishers Limited. All rights reserved©2013

Etot

Brown et al., Nature, November 2013

Deposition of Mass, Momentum & EnergyConservation of energy







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





• Given energy deposition we know the total energy released is area under profile

Etot

=

ZdE

dhdh ( + radiation )

area = Etot

14

020406080100Trajectory Distance (km)

20

40

60

80

100

Ener

gy D

epos

ition

(kt

/km

)

Deposition of Mass, Momentum & EnergyConservation of energy







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





• Given energy deposition we know the total energy released is area under profile

• For known entry angle, can rescale profile to be energy released along trajectory, x

Etot

=

ZdE


E

tot

=

ZdE

dx

dx

Etot

area = Etot

15


20

40

60

80

100

Ener

gy D

epos

ition

(kt

/km

)

Deposition of Mass, Momentum & EnergyConservation of energy• Given energy deposition we know the total

energy released is area under profile


Etot

=

ZdE


E

tot

=

ZdE

dx

dx

16

x

z ∆t

• This energy gets released into the mesh cells which intersect the tube surrounding the meteor at each time step, ∆t

Etot

energy deposited in ∆t


20

40

60

80

100

Ener

gy D

epos

ition

(kt

/km

)




Etot

=

ZdE


E

tot

=

ZdE

dx

dx

Etot

17







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





area = Etot


20

40

60

80

100

Ener

gy D

epos

ition

(kt

/km

)




• Since work = (force x distance), and aerodynamic drag is doing all the work, this profile is identically drag along the trajectory

Etot

=

ZdE


E

tot

=

ZdE

dx

dx

E

tot

=

ZdE

dx

dx =

ZD(x) dx A

erod

ynam

ic D

rag

(kt/k

m) Local value of

Drag, D (x )

18







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





area = Etot

• Mass loss equation

• Recall that aerodynamic drag is

• So mass loss is simply


dr

Conservation of mass & momentumdM

dt= ��CDSm

1

2⇢airU

3m


20

40

60

80

100

Ener

gy D

epos

ition

(kt

/km

)M

ass D

epos

ition

(kg/

km)

Local mass deposition

rate

19

dM

dt= ��DUm

D = CDSmq1 q1 =1

2⇢airU

2mwith

area = Total mass deposited




• Assuming constant Um and σ, local deposition of mass scales directly with Drag


dr


dt= ��CDSm

1

2⇢airU

3m


20

40

60

80

100

Ener

gy D

epos

ition

(kt

/km

)M

ass D

epos

ition

(kg/

km)


rate

20

dM

dt= ��DUm

D = CDSmq1 with


q1 =1

2⇢airU

2m




• Assuming constant Um and σ, local deposition of mass scales directly with Drag

• Area under profile is total mass deposited (Mentry – MGroundFragments)

• From mass deposition and velocity, we also know momentum deposition


dr


dt= ��CDSm

1

2⇢airU

3m


20

40

60

80

100

Ener

gy D

epos

ition

(kt

/km

)M

ass D

epos

ition

(kg/

km)

21

dM

dt= ��DUm

D = CDSmq1 with



rate

q1 =1

2⇢airU

2m

Overview


• Verification & Validation• Basic – Spherical charge examples• Chelyabinsk Case Study




22

Basic Verification & ValidationBlast from a spherical charge

• Static spherical charge with• No buoyancy• Etot = 520 kt, • Initial radius, ri = 1km

• Classical refs.• Brode, H. L.,Blast wave from a

spherical charge, J. Phys. Fluids. (1959) • D. L. Jones. Intermediate strength blast

wave. Physics of Fluids (1968)

2 4 6 8 10 12Distance (km)

-0.5

0

0.5

1

1.5

Ove

rpre

ssur

e (a

tm)

3.25 sec 6.46 sec10.24 sec14.03 sec19.85 sec

xri

Overpressure Evolution with Distance

ri = 1kmEtot = 520

Setup

23

Basic Verification & ValidationBlast from a spherical charge

Space-time overpressure evolution

• Etot = 520 kt, Initial radius, ri = 1km, no buoyancy

24

Basic Verification & ValidationBlasts over ground plane

• Numerous examples static and moving blasts over ground plane with buoyancy

• Static airburst with buoyancy

• Moving airburst

3DCart3DCart

t+100 t+200 t+500 t+1600

25




• Moving airburst

3DCart3DCart

t+100 t+200 t+500 t+1600

26


3DCart3DCart

t+100 t+200 t+500 t+1600



• Moving airburst

27

Validation: Chelyabinsk MeteorFebruary 15, 2013







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





Energy Deposition







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





Brightness

Very well studied event, simulations of virtually all aspects of entry, breakup, analysis of composition, blast propagation, ground damage, etc.

Brown et.al, Nature, 2013

• 12,500 metric tons, 19.8 m diameter

• Trajectory: • 18.6 km/sec, (~Mach 61.7)• 18° entry angle

• Data• Ground Damage (glass breakage data)• Shock arrival times• Light curve reconstruction• Energy deposition from infrasound measurements

28

Validation: Chelyabinsk MeteorFebruary 15, 2013• 12,500 metric tons, 19.8 m diameter

• Trajectory: • 18.6 km/sec, (~Mach 61.7)• 18° entry angle

• Data• Ground Damage (glass breakage data)• Shock arrival times• Light curve reconstruction• Energy deposition from infrasound measurements







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





Energy Deposition







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





Brightness

Brown et al, Nature, 2013• Primary references used• Popova & Jenniskens et al., Science Express, November 2013• Brown et al., Nature, November 2013• Chelyabinsk Airburst Consortium, + various other media

29

0 20 40 60 80 100Distance Along Trajectory (km)

10

20

30

40

50

60

Ener

gy D

epos

ition

(kt

/km

)

Validation: Chelyabinsk MeteorSimulation Details

New Time-Dependent chelyabinsk520kt

Cart3D

520 kt Chelyabinsk case with mass and momentum deposition

01020304050Distance (km)

0

10

20

30

40

50

60

Altit

ude

(km

)

0 10 20 30 40–100

18°

• Energy deposition:• Etot = (520 kt – 5% radiation)

• Profile from Brown et al. Nature 2013

• Net mass deposited: • mtot = 12.5e6 kg

• Trajectory: • 18.6 km/sec, (~Mach 61.7) @ 18° angle• Peak brightness @ 29.5 km• ~110 km length, 60→24 km altitude• Assume constant velocity

• 3D simulation with ~90M cells• Resolution of ~20 m along trajectory

& ~100 m resolution near ground• Simulation covers ~300 sec. of real time

Vertical Slice through trajectory

30

-40-30-20-100102030Distance (km)

0

10

20

30

40Al

titud

e (k

m)

-40-30-20-100102030Distance (km)

0

10

20

30

40

Altit

ude

(km

)-40-30-20-100102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

-40-30-20-100102030Distance (km)

0

10

20

30

40

Altit

ude

(km

)

-40-30-20-100102030Distance (km)

0

10

20

30

40

Altit

ude

(km

)

-40-30-20-100102030Distance (km)

0

10

20

30

40Al

titud

e (k

m)

t = 6.6 sec t = 6.6 sec

t = 30.3 sec t = 30.3 sec

t = 93.7 sec t = 93.7 sec

ln(P )� ln(P1)

ln(P1)

Local Mach Number Local Overpressure

31

Validation: Chelyabinsk Meteor

32

-30 -20 -10 0 10 20 30 40Downrange Distance, X, (km)

-30

-20

-10

0

10

20

30

Cros

sran

ge d

istan

ce, Y

, (km

)

Validation: Chelyabinsk MeteorGround footprint

• Goal is prediction of pressure & velocity on the ground

• Blast first contacts ground at t = ~82.7 sec elapsed time (~78 sec. after peak brightness)

• Excellent agreement with earliest data on blast arrival time data (76 – 90 sec) (Popova et al.)

-40-30-20-100102030Distance (km)

0

10

20

30

40

Altit

ude

(km

)

t = 93.7 sec

Vertical Slice through trajectory

Instantaneous Ground Pressure

t = 93.7 sec

33

Validation: Chelyabinsk MeteorGround footprint evolution

A’A

B’

B

A’A

B’B

t = 160 sec

Instantaneous Downrange Profile

Instantaneous Crossrange ProfilePeak Recorded Overpressure (so far...)

Instantaneous Pressure Profile

(inc. =1%)

34

Validation: Chelyabinsk MeteorGround footprint evolution

Instantaneous Downrange Profile

Instantaneous Crossrange ProfilePeak Recorded Overpressure (%)

Instantaneous Pressure Profile

(inc. =1%)

35

Validation: Chelyabinsk MeteorPeak Ground Overpressures

36

1%

2%

5%

6%

4%

3%


36

1%

2%

5%

6%

4%

3%

Glass Damage Data Comparison

61

Fig S38. Glass damage. Red data points were collected during the field survey, purple data were provided by the

Emergency Department. Open circles indicate that no glass damage occurred.

The most damaged settlements according to official data are shown by purple symbols in Fig.

S38. The damage area has an extent of about 180 km from north to south and 80 km from east to

west, but is shaped along a curved arc centered on Yemanzelinsk, extending from the northern

parts of Chelyabinsk as far south as Troitsk.

Red data points in Figure S38 are collected during the field survey, purple data points are

villages that reported damage through the Ministry of Emergencies at Chelyabinsk. Open circles

are sites where no damage occurred. The red points include many villages that had only a few

windows damaged (usually in school buildings). Hence, the inner contour of the purple points

may represent a higher overpressure than the outer contour of the red sites.

The value of overpressure, Δp, needed to break window glass is dependent on the glass

thickness and surface area. These values are not different between windows in Russia (most

affected buildings being from the 20th century) and other locations in the world. Glasstone and

Dolan [84] estimated the overpressure which caused essential glass damage at about Δp~3,500-

5,000 Pa. According to Mannan and Lees [87], an overpressure of about Δp~700 Pa is able to

Chelyabinsk 1%

2%3%

Y D

ista

nce

(km

)

X Distance (km)


• Glass damage data collected by the Chelyabinsk Airburst Consortium• Statistical correlation (Mannan & Lees) show 700 Pa (0.69%) shatters ~5% of

typical windows, 6% overpressure breaks roughly 90%.• Footprint similar to those in Popova et al. (ScienceExpress)• Breakage data estimate overpressure at chelyabinsk ~2-4% (P. Brown)

37

1%

2%

5%

6%

4%

3%

61

















Chelyabinsk 1%

2%3%

Y D

ista

nce

(km

)X Distance (km)

Validation: Chelyabinsk Meteor

38

Shock Arrival Time

Arrival time

t = 160 sec

Validation: Chelyabinsk MeteorShock Arrival Time

• Peak brightness at ~4 sec. elapsed time• First arrival at ~78 sec after peak brightness, • Predict ~90 sec (from peak brightness) at Korkino and Yemanzhelinsk• Arrival in vicinity of Chelyabinsk at 140 - 145 seconds• Neglected local wind, temperature and other effects of the real atmosphere• Overall very good agreement with data & best predictions in literature

39

contour ∆t = 33.2 secPredicted Shock Arrival Time (seconds)

282.1

83.1149

.4

215.8

61

















Chelyabinsk

contour ∆t = 33.2 sec

Shock Arrival Time (seconds)

282.1

83.1 149.4

215.8

Elapsed Time (sec) from 64km altitude

Overview

• Modeling & Solver





40

Overview

• Modeling & Solver





• Sensitivity to entry modeling

• Time-dependent compared to simple line source

• Entry angle / Spherical charge investigation

40

Sensitivity - Entry ModelingTime-Dependent modeling vs simple line source

• Entry only lasts for seconds, blast propagates for minutes

• Detailed entry modeling requires very fine time-scales (∆t ≈ 1.e-4)

• Cost: 90M cell simulation: (1000 cores x 8-12 hrs) - Under 1% of NAS Pleiades

• Line Source for mass, momentum and energy can reduce cost by 50%

• When is line source modeling appropriate?

41







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





Isobars

Line source along trajectory

Sensitivity - Entry Modeling

Time-Dependent

t* = 0.5482 (1.752 sec)

t* = 0.0082 (0.0272 sec)

t* = 2.0 (6.634 sec)

t* = 1.7 (5.63 sec)01020304050

Distance (km)

0

10

20

30

40

50

60

Altit

ude

(km

)

step 10001020304050

Distance (km)

0

10

20

30

40

50

60

Altit

ude

(km

)

step 1700


0

10

20

30

40

50

60

Altit

ude

(km

)


0

10

20

30

40

50

60

Altit

ude

(km

)

step 1000, ∆t = 0.001step 2000, ∆t = 0.001

Line Source

Seconds after entry

42


Time-Dependent

t* = 0.5482 (1.752 sec)

t* = 0.0082 (0.0272 sec)

t* = 2.0 (6.634 sec)

t* = 1.7 (5.63 sec)01020304050

Distance (km)

0

10

20

30

40

50

60

Altit

ude

(km

)

step 10001020304050

Distance (km)

0

10

20

30

40

50

60

Altit

ude

(km

)

step 1700


0

10

20

30

40

50

60

Altit

ude

(km

)


0

10

20

30

40

50

60

Altit

ude

(km

)

step 1000, ∆t = 0.001step 2000, ∆t = 0.001

Line Source

Seconds after entry

43

+ 1 sec + 1.7 sec

-40-30-20-100102030Distance (km)

0

10

20

30

40

Altit

ude

(km

)

-40-30-20-100102030Distance (km)

0

10

20

30

40

Altit

ude

(km

)

-40-30-20-100102030Distance (km)

0

10

20

30

40Al

titud

e (k

m)

-40-30-20-100102030Distance (km)

0

10

20

30

40

Altit

ude

(km

)

step 5520

step 7860 t* = 23.65 (78.45 sec)

t* = 10.15 (33.7 sec)step 7340

step 9670 t* = 24.938 (82.7 sec)

t* = 11.44 (37.9 sec)

Time-Dependent Line SourceMinutes after entry

44


Ground Footprint

• Some differences in highest overpressures (~1%) at earliest arrival time• Closer agreements at later time • Good agreement for location of peak ground pressure• Good agreement of arrival time from peak brightness• Geometry dictates that low-entry trajectories will show most discrepancy

45

Time-Dependent Line Source


Sensitivity - 90° Entry vs Spherical Charge

46

• Very cheap spherical charge models exist.

• Various handbook methods for damage estimation use spherical charge data

•Can these be used in risk assessment?• Where are these appropriate?• Perform quantitative assessment

• Investigate ground footprint• Accuracy of overpressure• Extent/strength of footprint• Details of impulse of blast on ground

How good is a static spherical charge model?

Sensitivity - 90° Entry vs Spherical Charge

47

Spherical ChargeLine-Source at 90°

• Etot = 520 kt, mtot = 12.5e6 kg

• Spherical charge @ 29.5km altitude, ri = 1km

• Line Source, 90° trajectory with same energy deposition as before

• Can also compare line source with 18° case for insight into entry angle sensitivity







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





How good is a static spherical charge model?

Sensitivity - 90° entry vs Spherical ChargeTime evolution


0

10

20

30

40

50

60

Altit

ude

(km

)01020304050

Distance (km)

0

10

20

30

40

50

60

Altit

ude

(km

)

t = 10.4 sec t = 10.2 sec

Line Source Spherical Charge

48

t = 10 sec

Sensitivity - 90° entry vs Spherical Charge

t = 19.6 sec0102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

Line Source

t = 19.6 sec0102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

Spherical ChargeMach contours

49

t = 19.6 sec

Time evolution


t = 19.6 sec0102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

Line Source

t = 19.6 sec0102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

Spherical ChargeMach contours

49







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





t = 19.6 sec

Time evolution


t = 44.5 sec0102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

Line Source


0

10

20

30

40

Altit

ude

(km

)

Spherical Charge

t = 44.5 sec

50

t = 44.5 sec

Time evolution


t = 118 sec0102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

Line Source

t = 118 sec0102030

Distance (km)

0

10

20

30

40

Altit

ude

(km

)

Spherical ChargeMach

contours

51

t = 118 sec

Time evolution

Sensitivity - 90° entry vs Spherical ChargeGround footprint comparison

-100 -80 -60 -40 -20 0 20 40 60 80 100Distance (km)

-4

-2

0

2

4

6

8

10

Ove

rpre

ssur

e (%

)

Line SourceSpherical ChargeEnvelope (Spherical Charge)

(overpressure)

(underpressure)

Envelope of Max and Min Ground PressuresSpherical ChargeLine Source







Time (s)–3.0 –2.0 –1.0 0.0 1.0 2.0 3.0

Abs

olut

e br

ight

ness

(mag

nitu

de)

–28

–26

–24

–22

–20

–18

a

Height (km)20253035404550

Ener

gy d

epos

ition

per

uni

t hei

ght

(kt k

m–1

)

0

20

40

60

80

100b





52

• Very similar envelopes - modulo details of energy deposition profile chosen.

• Both show lower peak overpressure than 18° trajectory


-100 -80 -60 -40 -20 0 20 40 60 80 100Distance (km)

-4

-2

0

2

4

6

8

10

Ove

rpre

ssur

e (%

)

Line SourceSpherical ChargeEnvelope (Spherical Charge)

(overpressure)

(underpressure)

Envelope of Max and Min Ground Pressures

Spherical Charge

Line Source

Surface Pressures @ t = 139 sec.

Surface Pressures @ t = 139 sec.

53

• Detailed pressure impulse shows similar instantaneous profiles as blast evolves

• Blast arrival times agree to within 2-3 seconds

Ground footprint comparison

Summary

• Outlined modeling for far-field propagation of airburst events using a Cartesian finite-volume method

• Showed basic verification and validation• Good prediction for model problems• Good prediction of footprint and arrival time data for Chelyabinsk meteor

• Showed envelopes and time-evolution of ground footprints for damage prediction and atmospheric-driven tsunami simulations

• Preliminary sensitivity investigations• Line source vs full time-dependent entry• Effects of Entry angle & comparison with specific spherical blast

54

Full paper for modeling and V&V planned at AIAA SciTech 2016 in San Diego (Jan 2016)

Atmospheric propagation and ground effects modeling

Next Steps...

• Parametric drivers• Vary entry angle, size and strength of asteroid• Parametric modification of energy deposition curve• Precompute parametric studies -- feed results to PRA

• Cratering & splashing

• Terrain and structures• Refine particular when scenario arises (e.g. PDC 15)

• Update models being output to Physics-Based Risk Analysis

• Update models being input from entry and breakup modeling

55

Atmospheric propagation and ground effects modeling

Thank You!• NASA Advanced Supercomputing Division – Task 3 & 4 teams

" Marian Nemec Donovan Mathias Jonathan Chiew George Anderson Chris Mattenberger Darrel Robertson

Lorien Wheeler

• Entry Systems Division – Task 2 team" " Dinesh Prabhu Ethiraj Venkatapathy

• New York University" " Marsha Berger

• NASA PD IPT" " James Arnold Craig Burkhard Jessie Dotson David Morrison Derek Sears

• IPT Seminar Speakers" " Peter Jenniskens Peter Brown Olga Popova Jay Melosh Paul Chodas

• NASA NEO Office" " Lindley Johnson

56

Far-Field Propagatin of Airburst Events Using a … Propagation of Airburst Events Using a Cartesian...

Documents

Transcript of Far-Field Propagatin of Airburst Events Using a … Propagation of Airburst Events Using a Cartesian...