NASA Technical Reports Server (NTRS) 20060047566: Overview Of Recent Enhancements To The Bumper-II Meteoroid and Orbital Debris Risk Assessment Tool

1.3 MB·English

by  NASA Technical Reports Server (NTRS)

IAC-06-B6.3.03

OVERVIEW OF RECENT ENHANCEMENTS TO THE BUMPER-II
METEOROID & ORBITAL DEBRIS RISK ASSESSMENT TOOL

Mr. James L. Hyde
Barrios Technology, United States, [email protected]

Dr. Eric L. Christiansen
NASA/Johnson Space Center, United States, [email protected]

Mr. Dana M. Lear
ERC, United States, [email protected]

Mr. Thomas G. Prior
Hamilton Sundstrand, United States, [email protected]

ABSTRACT
Discussion includes recent enhancements to the BUMPER-II program and input files in support of Shuttle
Return to Flight. Improvements to the mesh definitions of the finite element input model will be
presented. A BUMPER-II analysis process that was used to estimate statistical uncertainty is introduced.

INTRODUCTION 1 2
BUMPER-II is an MMOD risk analysis program
originally developed by for the Space Station
Freedom Program. Over the years, the
capabilities of this engineering analysis tool
have been extended to include the Space Shuttle
Orbiter, the International Space Station (ISS)
and many other spacecraft. When provided with 3 4
a vehicle shape, orbit parameters and applicable
ballistic limit equations with defined failure
criteria, the BUMPER-II code will calculate the
MMOD risk for spacecraft in low Earth orbit
against a variety of natural and man-made
environments. Thousands of hypervelocity
impact tests have been performed on Fig. 1: BUMPER-II functional overview.
representative samples of ISS shields and
subsystems, Shuttle thermal protection system FAILURE CRITERIA UPDATES
(TPS) materials [1], Extravehicular Mobility As part of the Shuttle Return-to-Flight effort, the
Unit (EMU) materials [2] and other spacecraft NASA Johnson Space Center’s Hypervelocity
components to determine MMOD impact Impact Technology Facility performed
parameters at the failure limits of the various hypervelocity impact testing and analysis of
subsystems. BUMPER is used to calculate Shuttle wing leading-edge (WLE) reinforced
MMOD impact risks to specific Orbiter surfaces. carbon-carbon (RCC) test samples to update
An integrated mission assessment is completed WLE threshold failure criteria [4]. After the
using Poisson statistics and knowledge of the hypervelocity impact tests, the samples were
distribution of times spent in each unique exposed to typical reentry heating conditions at
Orbiter attitude [3]. the NASA JSC Arc-Jet (AJ) Facility to
determine the extent of heating induced damage
growth.
RCC test article after HVI test

Fig. 3: Pre-STS-107 RCC Failure Criteria Map.

RCC test article after HVI & AJ testing

Fig. 4: Post-STS-107 RCC Failure Criteria Map.

Fig. 2: 0.8 mm aluminum HVI/AJ RCC test results. ORBITER/ISS MATED ATTITUDES
The Shuttle and International Space Station
It was found from the HVI/AJ testing that non- (ISS) Programs have decreased MMOD impact
penetrating pits would lead to burn-through in risks to the orbiter by changing the orientation of
some areas of the WLE where burn-through can the ISS while the shuttle is docked. The change
lead to loss-of-vehicle (LOV) during reentry. in orientation – essentially flying the ISS
Figure 2 shows the resulting damage caused by a “backwards” – provided incidental shielding to
0.8 mm diameter aluminum hypervelocity the shuttle as well as directing MMOD sensitive
impactor and subsequent damage growth due to areas of the WLE and nose cap away from the
AJ testing. For STS-107 and previous missions, majority of the MMOD particle flux. Figure 5
WLE failure threshold consisted of 1 inch shows the shuttle-ISS docked orientation change
diameter allowable hole sizes in RCC on the with respect to the ISS velocity direction. In
upper surface and ¼ in. hole size on the lower previous ISS missions, the belly of the vehicle
surface. The results of the recent RCC/AJ testing typically faced into the velocity direction of ISS
indicated that the WLE failure criteria for LOV motion and highest MMOD impact flux. The
should be reduced for MMOD assessments on attitude change orients the bottom of the shuttle
future missions. Figures 3 and 4 show the WLE in the wake direction of ISS reducing MMOD
and nose failure criteria maps before and after impacts to the most vulnerable surfaces of the
the recent changes. The reduction in allowable vehicle and improving crew safety and odds of
damage results in increased calculated MMOD mission success.
risks for future missions [5].
vveelloocciittyy vveelloocciittyy
ISS –XVV mated attitude N ISS +XVV mated attitude
Odds of M/OD Critical Penetration Earth Odds of M/OD Critical Penetration
1 in 151 1 in 78
Vehicle Orientation
Fig. 5: Shuttle-ISS mated attitudes before STS-107 (ISS +XVV) & after STS-107 (ISS -XVV).

A trade study was performed to assess the the odds of MMOD critical penetration from “1
quantitative effect of the ISS -XVV attitude in 78” to “1 in 151”, an overall mission risk
change on the critical MMOD risk for the STS- reduction of nearly 2X. The bar chart in figure 6
114 mission. The results of the mated attitude illustrates the qualitative effect of orbiter attitude
trade study are illustrated in figure 5. The ISS - on MMOD critical penetration risk using the
XVV mated attitude used for STS-114 increased RCC failure criteria shown in figure 4.

MM//OODD CCrriittiiccaall PPeenneettrraattiioonn RRiisskk vvss.. SSppaaccee SShhuuttttllee OOrriieennttaattiioonn
CCooddee:: BBuummppeerr--IIII vv22..3311ff ------ AAllttiittuuddee:: 221166 nnmm ------ IInncclliinnaattiioonn:: 5511..66oo ------ FFlliigghhtt YYeeaarr:: 22000066
DDeebbrriiss EEnnvviirroonnmmeenntt:: OORRDDEEMM22kk ------ DDeebbrriiss DDeennssiittyy:: 22..88gg//cccc ------ MMeetteeoorrooiidd EEnnvviirroonnmmeenntt:: SSSSPP3300442255 ------ MMeetteeoorrooiidd DDeennssiittyy:: vvaarriiaabbllee
33..00%% PPrree SSTTSS--110077
NN
22..55%% EEaarrtthh
VVeehhiiccllee OOrriieennttaattiioonn
%% 22..00%% 11 iinn 5500
ation Risk ation Risk ds = 1 in "x")ds = 1 in "x")
ay Penetray Penetrnetration Odnetration Od 11..55%%
10 D10 D(Pe(Pe 11..00%% 11 iinn 110000 PPoosstt SSTTSS--110077
11 iinn 220000
00..55%%
%% %% %% %% %% %% %% %% %% %% %% %% %% %% %% %%
33 44 44 44 55 88 00 11 44 44 00 11 22 44 55 88
00..00%% 0.0. 0.0. 0.0. 0.0. 0.0. 0.0. 1.1. 1.1. 1.1. 1.1. 2.2. 2.2. 2.2. 2.2. 2.2. 2.2.
++ZZLLVV ++XXLLVV --XXLLVV --ZZLLVV ±±YYLLVV ++ZZLLVV --ZZLLVV ++ZZLLVV --ZZLLVV ±±YYLLVV ++XXLLVV --XXLLVV ++XXLLVV --XXLLVV ±±YYLLVV ±±YYLLVV
--XXVVVV --ZZVVVV --ZZVVVV --XXVVVV --ZZVVVV ±±YYVVVV ±±YYVVVV ++XXVVVV ++XXVVVV --XXVVVV ±±YYVVVV ±±YYVVVV ++ZZVVVV ++ZZVVVV ++ZZVVVV ++XXVVVV

Fig. 6: Critical MMOD risk as a function of orbiter attitude.
5000 elements. These newly revised mesh areas
provide a significant increase in analytical detail,
allowing property definition and risk
calculations for specific regions of individual
WLE components such as panels and seals.
Figure 8 depicts the new finite element mesh in
the nose cap/chin panel region of the Shuttle.

PPrree SSTTSS--110077
PPrree SSTTSS--110077
PPoosstt SSTTSS--110077

Fig. 7: WLE RCC finite element mesh.

ORBITER FEM UPDATES
Figure 7 shows a detail of the WLE RCC area of
the orbiter that is represented by a detailed finite
element mesh containing over 50,000 elements. PPoosstt SSTTSS--110077

Each color change represents a different failure
Fig. 8: Nose cap/chin panel finite element mesh.
criteria region. This mesh area has 592 distinct

regions, where hypervelocity impact damage

resistance is defined to reflect failure criteria and

physical differences such as location, thickness,

and material. [6]

The Nose Cap/Chin Panel area of the orbiter was

also revised and is now composed of nearly
ESTIMATION OF UNCERTAINTY BOUNDS 4.5
4.0
Teshtiism saetec tiounn cdeorctauimnteyn tsb oa upnrdosp oisne d BteUchMnPiqEuRe -tIoI BIDBIDnnUipUispsMutMutrrtPitP ib bEEuuRRttii o onn Values x 10^2112233......050505
assessments of MMOD risk to the shuttle. A DDeeffiinniittiioonnss 00..050.9890.9900.9910.9920.9930.9940.9950.9960.9970.9980.9991.000
typical analysis is comprised of input scripts and 5.0% 9 0.0%5.0%
0.994360.99763
data files for the Bumper program modules BBUUMMPPEERR iinnppuutt
ppaarraammeetteerr sseett
(Geometry, Response & Shield) that describe the
input parameters for the mission. This EExxtteerrnnaall SSttaattiissttiiccaall PPaacckkaaggee
architecture allows for multiple analysis cases to
BBUUMMPPEERR rruunnss
set up in a script through loop constructs. [7] ((mmuullttiippllee))

SSttaattiissttiiccaall AAnnaallyyssiiss ooff RRiisskk RReessuullttss && UUnncceerrttaaiinnttiieess
Methodology
11 iinn 770000
The BUMPER-II code calculates an expected 11 iinn 660000 NNoottiioonnaall OOuuttppuutt//UUnncceerrttaaiinnttyy AAnnaallyyssiiss
number of penetrations of a spacecraft based on 11 iinn 550011 00iinn 550000
a single set of program inputs. To estimate D RiskD Risk 11 iinn 33001111 00iiiinnnn 343400000000
uncertainty bounds, multiple instances of the MOMO 11 iinn 220000
same calculation need to be run while varying MM 11 iinn 110011 00iinn 110000
11 iinn 0000
selected environment and penetration input 11 55 99113311772211BB2255UU2299MM3333PP33EE77RR4411 RR4455uu4499nn55 33RR55ee77ss6611uull6655ttss66997733777788118855889999339977
parameters. The “environment” variables for this Fig. 9: Risk assessment workflow.
study were MMOD flux and density, OD
velocity distribution. Ballistic limit equations for each case in the parameter set. MMOD risk
and failure criteria were selected as the output values from the runs is tabulated and sent
“penetration” variables. Figure 9 provides an back to the statistics processor, where
outline of the risk assessment methodology used uncertainty bounds can be determined. The
in this assessment. The upper “BUMPER Input” technique used in this analysis for the estimation
boxes describe the input phase for the analysis; of uncertainty bounds incorporates Bumper code
the intermediate boxes indicate calculations and input variables from the sensitivity analysis with
the lower box represents output processing. the added complexity of allowing values in
The first step in the analysis is the statistical between the upper and lower bounds. Input
modeling of selected input parameters. The input distribution types chosen for each variable are
distribution definitions are used as directives to provided in table 1.
an external (i.e., not part of the Bumper code)
statistical processor that generates randomly Probability Distribution Functions (PDF)
selected values of Bumper input parameters. The A “discrete” distribution selects from a defined
input parameter set is then copied into a set of inputs while a “continuous” distribution
BUMPER-II input file which runs the code once can return any value between the limits. The
uncertainty analysis uses a statistical
preprocessor (@Risk v4.5) to select values for

Distribution
Type Name Type PDF Limit Values
Environment OD Flux Continuous Log Normal, μ=1.0 0.9 1.6
Environment MM Flux Continuous Log Normal, μ=1.0 0.5 2.0
Environment OD Velocity Discrete --- Min, Base, Max
Environment OD Density Continuous Log Normal, μ=2.8 1.0 7.9
Environment MM Denisty Continuous Log Normal, μ=1.0 0.5 1.9
Penetration MM/OD BLE Continuous Extreme Value, μ=1.0 0.9 1.15
Penetration MMOD Failure Criteria Discrete --- Min, Base, Max

Table 1: Input distribution types.
each of the Bumper inputs using the probability
UUnnmmaatteedd AAttttiittuuddee ––RRPPYY==00°°,,118800°°,,00°°
distribution functions (PDF) discussed below.
NNoossee AAfftt,, BBaayy EEaarrtthh ((33..112255 DDaayyss))
The Log-normal PDF for OD and MM flux
input are the only PDF’s with uncertainty values
of 68% (±1 sigma) at the distribution limits.
The other continuous distributions in this
assessment were assumed to have 95%
VVVVVV
confidence bounds between parameter value
extrema. MM and OD density Log Normal input
distributions were modeled with 95% confidence
bounds at the extrema and the mean set to the LLLVVV
baseline analysis value (OD density = 2.8 g/cc
and MM density = 1.0 g/cc). 11.. OOrrbbiitt IInnsseerrtt (cid:198)(cid:198)DDoocckk
33.. UUnnddoocckk (cid:198)(cid:198)RRee--eennttrryy
The discrete PDF for the OD velocity
distribution was set to select the nominal
MMaatteedd AAttttiittuuddee ––RRPPYY==00°°,,111133°°,,00°°
ORDEM2000 orbital debris velocity distribution
NNoossee SSppaaccee,, BBaayy AAfftt((77..662255 DDaayyss))
option twice as often as the equally likely
minimum and maximum velocity distribution
options.
The Penetration PDF’s used in the uncertainty
analysis are correlated to shield penetration, so
they were not modeled as separate distributions
for meteoroids and orbital debris. An Extreme
Value PDF was used to model the Ballistic
Limit Equation (BLE) factor inputs, with a mean
at the nominal analysis baseline of 1.0 and ~95%
of the area between the analysis extrema of 0.9
and 1.15. Three discrete options are modeled for
failure criteria. The “max” and “min” options
were assumed to be half as likely to occur as the
“nominal” option.
VVVVVV
Assessment Parameters
The following constants were used for the
uncertainty calculations in BUMPER-II:
LLLVVV
OD Environnent – ORDEM 2000
MM Environnent – SSP-30425, Rev. B
Year – 2003
Inclination – 51.6° 22.. DDoocckk(cid:198)(cid:198)UUnnddoocckk
Altitude – 400km
Fig. 10: BUMPER-II finite element models.

Unmated
Analysis Results
Attitude – RPY = 0°, 180°, 0°
Figure 11 shows the frequency plot of the
Exposure Time – 3.125 days (29%)
MMOD critical risk output for the 180 cases in

this analysis. Using @Risk, a probability
Mated
distribution was fitted the output values. Figure
Attitude – RPY = 0°, 113°, 0°
12 shows the 90% uncertainty bounds on the
Exposure Time – 7.625 days (71%)
example baseline analysis of 1 in 440 vary from

a minimum of “1 in 114” to “1 in 2654”.
Figure 10 depicts the two finite element models

that were used for the uncertainty assessment.
the uncertainty assessment based on the new
80
failure criteria as well as reflecting the latest
70
changes in uncertainty ranges of key risk
60 assessment parameters Additional research into
the parameters used in the input PDF’s may also
50
Frequency40 b e desirable.
30 REFERENCES

20
1. Christiansen E.L. and Friesen L.J., “Penetration
10
Equations for Thermal Protection Materials”, Int.
0 Journal Impact Engineering, Volume 20, 1997.
38 439 841 1242 1643 2045 2446 2847 3249 3650 4051 4453 4854 More
Bin
2. Christiansen E.L. Cour-Palais B.G. and Friesen
Fig. 11: MMOD risk frequency plot.
L.J., “Extravehicular Activity Suit Penetration

Resistance”, Int. Journal Impact Engineering,

Bumper Uncertainty Analysis - Output Distribution Volume 23, 1999.
1.4
3. Hyde J.L. and Christiansen E.L., “Space Shuttle
Mean = 851.02
1.2 Meteoroid & Orbital Debris Threat Assessment
Mean = 851.02 Handbook: Using the BUMPER-II Code for
1.0 Shuttle Analysis”, JSC-29581, December 2001.
Values x 10^-3 00..68 InputFit 4. LfFolyrigo MhnstM” F, O.J,S D“CR I-Cm62Cp9a H2c6ty ,Ap Mesrsaveresclshom c2ei0tny0t 5IPm. aprta 1ct, TReestutirnng t o

0.4 5. Lyons F., “RCC Hypervelocity Impact Testing
for MMOD Impact Assessment Part 2, STS-114
0.2 Flight Readiness Review”, JSC-62988, May
2005.
0.0
Baseline Od0ds(cid:206)1 in 440 1000 MMO2000D Critical Pene3000tration Odds 4000 5000 6000 6. Lear D., et al., “Enhancement of Space Shuttle
< 90.0% 5.0% > Models Used in BUMPER Micrometeoroid and
114 2654
Orbital Debris Risk Analyses”, Orbital Debris
Fig. 12: MMOD risk distribution. Quarterly News, Volume 9 Issue 2, April 2005.

Forward Work 7. Hyde J.L. and Christiansen E.L., “Bumper-II
The next version of the uncertainty analysis will Meteoroid and Orbital Debris Threat Assessment
support the updates to the orbiter finite element Code: Sensitivity Analysis & Uncertainty
Bounds”, JSC-63313, September 2005.
model documented in this paper. The failure

criteria changes to the RCC wing leading edge
and the nose cap areas significantly effects
MMOD risk. Forward work includes updating