NASA Technical Memorandum 107266

Cassini/Titan IV Acoustic Blanket

Development and Testing

William O. Hughes and Anne M. McNelisLewis Research Center

Cleveland, Ohio

Prepared for the42nd Annual Technical Meeting and Exposition

sponsored by the Institute of Environmental SciencesOrlando, Florida, May 12-16, 1996

National Aeronautics and

Space Administration

https://ntrs.nasa.gov/search.jsp?R=19960047134 2018-05-27T09:13:31+00:00Z

_AN IV ACOUSTIC BI,ANKEr DEVELOPMENT AND _G

Wmi_m O. Hughes, NASA Lewis Research CenterAnne M. McNelis, NASA Lewis Research Center

BIOGRAPHY

William O. Hughes develops and directs the launchvehicle v_roacoustic environment activities at NASALewis Research Center. He has defined acoustics and

v_ration requiremmts, specifications and test plans forover 10 years. Mr. Hughes previously worked atAnalex Corporation, U.S. Steel Research and Raytheon.He received his B_. (1977) in Physics from Penn StateUnivemity and his ME. (1981) in MechanicalEngineering from rneg_Mellon Univcmity.

Anne M. McNdiJ is a s_ dynamics engineer atNASA I_zwis Research Center. For the past three years,

she has performed vibroacousfic analysis to developacoustic and random v_ration test specifications. Ms.

McNelLs previously worked for Bailey ControlsCompany. Her B.S. degree is in Systems and Contro!Eng0neering fzom Case Western Reserve University.

_CT

NASA Lewis Research Center recently led a multi-

ot_nizgio_ effort to develop and test verify newacoustic blankets. These blankets support NASA's goalin reducing the _tan IV payload faLrmginternalacoustic environment to allowable levels for tl_ Cassini

spacecraft. To accomplish this goat a two phaseacoustic test program was utilized. Phase One comistedof testing numerous blanket designs in a fiat panelconfiguration. Phase Two consisted of testing the mostpromising designs out of Phase One in a full scalecylindrical payload fairing. This paper will summarizethis highly successful test program by providing therationale and resulLsfor each test phase, the impacts of

this testing on the Cassini mission, as well as lxovidingsome general informationon blanket designs.

KEYWORBS

Acoustics, Acoustic Blankets, Blankets, Cassini, Payload

Fairing, Spaoect_ Acoustic Environment, SpacecraftAcoustic Testing, Titan IV, Vibroacoustics

INTRODUCTION

New and improved acoustic blankets were recentlydeveloped and tested to support NASA's Cassinimission. Acoustic blankets are utilized in the payload

fairing (PLF) of expendable launch vehicles (ELVs) toreduce the fairing's interior acoustics and the subsequentvibration resp(msc of the s_ and its components.

The CX im spacea wm be Launchedin Octa 1997, by a Titan Iv/centaur launch vehicle, to exploreSaturn and its moons. "[he electric power source for theCassini mission are three mission cri_cal RadioisotopeThermoelectric Gcncratom(RTGs). The RTG design

wm previously vibration qualified for the Space Shuttle_t,mch cDvirotmactRand utilized on _c C.ralJlco

Ulysses spacecraft missions.

However analysts at the JetPropulsion Lalx)ratoryOPL), the spacecraft designer, preOiaed that acoustimllydriven vibration levels for the Cassini RTGs wouldexceed the RTGs' previous qualification v_ration levels.This cxceedence is primarily due to RTG mountingdifferences along with differences in the hunch vehicle

andspacecra .

To avoid an extremely costly reqtmlification of theRTGs, a major acoustic blanket development and testeffort was initiated and funded by NASA Lewis

Research Center (LeRC), the latmch vehicle integratorfor the Cassini mission. If successful the new acoustic

blankets would provide a lower acoustic and vibrationenvironment for the Cassini's RTGs than the

environment obtained when using the standard Titan IVacoustic blankets.

BesidesNASA LeRC and JPL, other organizationsinvolved in this joint effort included Lockheed MartinAstronautics (LMA, formerly Martin MariettaTectmologies Incorporated, MMTI), McDonneU DouglasAer_pace (ME)A),_ Corporation,Analex

Corporation,CambridgeCollaborafiveIncorporatedandtheRiverbankAcxmsticalI.zboratory(RAL).

ACOUSTIC 'lEST PROGRAM OVERVIEW

A_ousticblanbt clcsigntcc_zology for aerospaceapplicationshas seen tittle clcvclopmcntin the pasttwenty-five years. DeveJ.oping an acoustic blanket tomeet the needs of the Cassini mission necessitated

developing advanced blanket tedmology. Not only didthe blanket have to reduce the acousticfield

sLon_fu'_ntly, but it bad to do so in thefrequency range of 200 to 250 Hz. Typicany acousticblankets are most effective at frequencies of 400 Hz andabove.

Specifically, our goal was m design and test a newacoustic blanket which woukl reduce the expectedacoustic envirmm=t for the Cassini RTGs by 3 dB at200 and 250 Hz, when compared with the baselim Titan

IV blanket system environment.

"Ihe apFoach taken was to develop a two-phase acoustictest program that would provide confidence that the newblanket would x=ult in an optimal, feasible system thathad a high probability of performing we. in the flightco_figuration.

Phase One consisted of evaluating new blanket designsby acoustic testing of fiat panel blanket samples. Flatpanel testing had the advantage that numerous designscould be quickly evaluated at a relatively low cost. Byproper interpmati_ of the absorpti¢_ and transmissionloss test data obtained, the leading candidate designs¢xmld then be ch=en for further testing in Phase Two.

Phase Two would test the leading candidate blanketdesigns=ci the baseline Titan IV blanket design in a_dl scale cylindrical payload fairing. A]though this typeof testing is expensive, the effect of the blankets onredudng the PI.Fs interior acoustics would be measuredwiththe ight-like andgeometry, for oaly the few lxom_ng candidates.

This two phase t=t approach was dx=a beca=e it wasozsid=ed _lc and risky to test m unprovennew design in an expensive full scale test. Likewise thegeometry and size of the flight payload fairing made itunwise to baseblanketselectionsolely on the basisoftesting of flat panel samples.

Toe two phase test program which was foUowed isilh_trated in flow_ format in F'_qa'e 1.

Verification Conliguralions(including some outside of

Design ol Experiments Limits)

IYes

Ied esl9 n Posslbts ?_(__

rner weight, etc) I [J

!No

Stop:

No BJanket

Redes gn

Acoustic Panel Tests at

Riverbank Acoustical Laboratory

(parameter opUmizallon on barrier

rnatedal, location, weighl, etc.)

I9 Configurations Tested

Design of Experiments Techniques Used

Quamtlfy _ Effect ¢4' Blankets "_

easure Basic Blanket Properties (TI., all.a) I

Calculate A Effect of Blankets ,,/

+Results Promising? (A. > 2 dB)

IYes vl0 - 5- _ Blanlm

!

Casslnl Payload Fairing Acoustic Tesls [ Team Effort led by

J Marlin Mariella7 full-scale Acoustic Tests al I .Jan 10 - Feb 14. 1995

Martin Marietta's Acoustic Chamber IV10 and V5

Compared tO Baseline ConfigurationA Effects Quantified

Both V5 and Vl0 Blankets _

el Design Goal of 3 dB Acoustic I

Reduction at 200 & 250 Hz .J

Team Eflort led byl_IcOonnellDouglas

March 28 - April 29. 1994

PHASE ONE

PHASE TWO

lqliure 1. Overdew ottbeTwo PhaseAcousdc Blanket Test _

FLAT PANEL TESTING

DEVELOPMENTAL SERIES OVERVIEW

The testing of the new blanket designs in a f_t panel

configuration ocaLrred in March-April 1994 at theRivaba_ Acousticm Laboratory(RAL), Geneva,Hlinois. Absorpdon values for the blankets wereobtained from reverberation time tests per ASTM CA23.

Blanket transmission loss(TL) values were obtained

from testing per ASTM Eg0. Figure 2 illustrates the TLtest configuration. Umi_ng the absorption and "IT,test

data, analytical predictions were made to calculate the

effect of each new blanket design in redudng the PLF's

intmor acoustics, at the frequencies of interest.

_ MLLI_ WJUJ.

PAN m

t IH_4_KH_rS NO1' -- IIEVl[R_IU4ilrIIOUmCE SlOE

/"A total of 19 different blankets (18 new designs and the

Titan IV baseline blanket) were tested for absorption mw_ --,e"_m'-"

and TL characteristics. "Fnese designs are illustrated in

F_n'e 3. Additionally, a isogrid panel sample, fi'om aTitan IV PLF wall, had its TL measute_ separately and Figure 2.

was also used for all the blanket TL testing. (The Titan

IV PLY is a cylindrical aluminum isogrid structure,

mnsisting of a geometric pattern of machined out

triangular pockets.) Each blanket tested was an 8 foot

by 9 foot _aangular sample. As a material constraint,all blanket materials utilized in the new designs had to

be already qualified for spaceflight.

Testing was divided into two series of tests known mthe development tcsts and the verification tests. The

development test series will be explained first. As part

of the development tests, the isogrid panel and the TitanIV baseline blanket wcrc tested. Also the Design of

Expcdmcnts (DOE) tcctmiquc was utilized to maximizethe amount of meaningful test information while runninga minimum number of tests. From this DOE technique

it w_ cxpected that one could determine the infl_noe

of various factors on the response and determine whichcombination of these factors would optimize the

response. Each development testwas run twiceto

check forreasonablemeasurement repeatabilityand

insurethatthetestwas recordingmeaningfuldataand

not just backgn3und variation.

'I'ne Titan IV baseline blanket is 3 inches thick, with a

0.6 pounds per cubic foot deasity fiberglass batting,with no internal barrier. It was believed that blanket

improvements cadd be obtained by optimizing both the

absorption(i.e.thickerblanket)and TL (i.e.heavier

blanket) characteristics forour Cassinimission critical

frequencies of 200 - 250 Hz. Therefore in order toreach our acoustic goal either a thicker (four inches)blanket and/or a blanketwith an internalbarrier would

be needed.

Transmission _ Test Configuration

for Flat Paad Testing (Chamber walls

are cut away to show isogrid panel and

[mer mm) (Fr_ mterenee L)

The DOE part of the development tests looked at three

main parameters of a 4 inch thick blanket design. Fust,

the density of the blanket's fiberglass batting was variedfrom 0.6 to 2.4 pounds per cubic foot (pcf). Second,

the density of the internal barrier was varied from 0.0

(no internal barrier) to 0.44 pounds per square foot

(pso. Third,the locationof the internalbarrierwasvaried from 0 to 3 inches fi'om the isogrid paneL

FLAT PANEL TESTING

DEVELOPMENTAL SERIES RESULTS

The results from the devck>pmcntal test series were

surprising and somcwlmt disappointing. With regards to

the absorption data it was discovered that the Titan IVbaseline blanket was already optimized for our

frequcndcs of interest. The absorption peak of theTitan IV baseline blanketwas previously thought to

occm" between 400 and 5(]0 Hz. The flat panel test of

the baseline blanket showed this peak to be at 250 Hz.

Increasing the blanket(batting) thickness, such as in

DOE 1, improved the absorptionat 125 Hz, but aauaUymade the absorption worse at 200 - 250 Hz, as shown

by Figure 4. Thus our new intent was to try to keepour baseline blanketabsorption values and toreach our

goal by increasing the "IT, at 250 I-Iz.

hmrme

0.6

0.6D

0.6

_;,;[ ........

DOE 1

0.6

0.6

0.8

0.6

XXI(XXXXX3Lt(X

DOE 2

1.2

1.2

1.2

1.2

xlooixxxxxxx

DOE 3

1.2

i.2

2.4

2.4

x --- ;, .......

OOE 4

0.6

0.6 .24 lid

0.6

0.6

Xlv_rlr_vwvv

DOE •

!.2

1.2

1.2 .44

1.2

Vl

DOE S

1.2 .24 psf

1.2

1.2

1.2

X_vvvvv,rm_1

_E9

1.2

1.2 ,44. ast

2.4

2.4

x.loo(xxxxxxx

DOE 6

1.2

1.2

2,4 .24 Ds/

2.4

V1

0.6

O.S .44 m;f

0.6

0.6

_r6

0.6

0.6

0.6

X lr,vv,rv_m,tn_r

V2

0.6

0.6 .44 n_

2.4

2.4

0.6

0.6

0.6D

O.6 .44 us/

v ...........

V4

0.6

0.6

0.6 .44 m;f

V10

0.6 0.6

0.6 0.6

0.6 .44 md 0.6 .88 tort

0.6 ! .2

(3" 0ap! 0.6 1.2

0.6 zu_uu,,,u_

XXXXXXXXXXXX XXXXX XXXXXX

Vll

o.6

o.6

0.6

o.6 .88 psi1.2

1.2

V12

0.6

0.6

0.6 .88 Nf

1.2

x_

V13 --

0*6 --

0.6

0.6

IJBGEI_:-- ]Skai_ l)edp

_UInS l)_¢q (pd)f _rrk," ])em_ (!_

.sspsf _ _rrkrl_j_"

_.z I-- (mob 1 lacb_1.2

_, -- bqFtd i'md

]Rigure 3. Blanket Destgm Tested in Phase One

4

To maintain this baseline absorption required a

minimum of 3 inches of fiberglass ba_ng on the

inboard side (side opposite PLF isogrid wall) of the

barrier. The presence of an intemat _n'ier not onlyincreases the TL but itaffectsthe blanketabsorlXio_

characteristics by creating a double peak (and a valley

between the peaks) in the absorption spectrum. "t_nus it

also became important to avoid sttifting the absorption

valley into the critical frequency range of interest.

Hgurc 4 illustrates these points.

With regards to TL, it was found that the barrier neededto be either heavier or plac_l further away fxum the

PLF isogrid walL Analysis of the test data showed thata 4 inch blanket would not meet our goats. Both the

_tion and TL requireds_n_c_t tlncknc_ andtherefore even thicker blankets would be nccdcd to

reach our goal

FLAT PANEL _G

VERIFICATION SERIES OVERVIEW

The original intent of the verification series was to testverify the optimum blanket candidates as identified by

the DOE technique. Although much useful information

cbtzined in the developmentalseries, there was no

design tested which met our goals nor did the DOE

analysis point to any combination of the tested blanket

parameters which would meet our goals.The name"verificationseries" remained but thisseries now

became an effort to use the previous test data to

analytically brainstorm to a solution within the allowable

budget, blanket materials and test facility timeconstraints.

A few verification blanket designs were tested with

mixed, but non-satisfactory, results. As indicated earliera thicker blanket would be needed. Relief came from

the LMA Cassini Project Office who indicated that a 5inch thick and even a 6 inch thick blanket would be

allowable and stilt meet the necessary mission clearance

reqttirements.

12 -

1 •

'_M08 •

g

_Q'06o

0d

Figure 4.

Absorption Flat Panel Test Data

02

!0 I I I I I I I I I I I I I I I I I I I

Frequency (Hz)

BASELINE _ DOE 1 _DOE9

14 -

]_llll'e$.

Al_orplion Flat Panel Test Data

12 -

i -

| .08 -

g .

o

04 •

0.2"

o I I I I I I I I I I I I I I I I I I I

Frequency (!._)

[ -,,=_ BASELINE "'4---- vs _ vto I

Figure 6.Transmission Loss Flat Panel Test Data

70 -

60 -

20,

IO

I I I I I I I I I I I I I I I I I I I

Frequency (Hz)

=--,7--- BASELINE _ ISOGRID ""4---" V5 -.--o-.-- VIO

6

FLAT PANEL TESTING

VERIFICATION SERIES RESULTS

Configuration 3/5 was the first 6 inch thick bla_cttested. V5 also had a heavy internal bamer (0.44 psi)

and was the first new blankct design which stxJwed

significant promise in meeting the original test goaL

C_xx_guration V10 was the first test configuration to

have the "super" heavy barrier (0.88 _ and was 5

inches thick. ARhough this blanket would weigh about6 timesthe Titan IV baselineblanket,thisweightwas

allowedby thc Cassiniprogram.

Thinner (4 inches)and thicker(6inchcs)varimionsof

thesuper heavy barrierwere alsotestedin

configurationsV12 and Vll, V13 respectively.

From all19 new blanlmtconfigurationstemed,VIO,

Vll and V13 configurations,allwith the superheavy

barrier,were analyzedtoreducethe PLF's acousticthe

best at 200 - 250 Hz. V5 was the only configuration

without the super heavy bamer which was analyzed to

o.r goaL

Figur_ 5 illmtra_ the fiat panel absccption test data forthe Titan IV baseline,, V5 and Vl0 blankets. Similarly,

F'_n-e 6 illustrates the TL test data for these same

blankets and the isogrid pan©l by itself.

It should be noted that the impact of the increased "IT,

values seen in the fiat panel test rcsulls is scvgndy

lessened when prediaing PLF ao0ustic noise reduction.

This is because the flight PLF does not have 100% full

blanketcoverage,but isinsteadonly partiallycoveredtoallowforaccess,doocs,splitrails,wiringharnesses,etc.

Using the measured _ and TL flat paneltest

data,MDA performed acousticanalysisusingtheir

PLFNOISE so_ to ixedia the noise reduakmwhich would be obtained for the T'R_ IV PLF with the

appropriate flight blanketcoverage. Based on this

analysis,itwas decidedtochoose VlO as theleading

blanket candidate fex furth¢_ testing in the full scale

PLF oonfiguration. V5 was also chosen for this

additional testing because its rmulls also looked

promising and itsdesign(barrierweight)was

significantly different Rom VlO.

A complae summary of the flat panel test results maybe found in Rcfetctm_ 1.

FULL SCALE I'LF _G OVERVIEW

Having chosen the most promising blanket candidates

(Vl0 and VS) out of Phase One testing, Phase Two

testing could now begin. Phase Two testing was a funscale test series with a cylindrical PLF, which would

simulate the flight boundary ox_litions and geometry.

Phase Two testingoccurredinJanuary-February1995 at

LMA's Rot¢_'ant Acoustic Laboratory (RAL),

Dearer, Colorado.

The test hardware consisted of a 60 foot high section of

a Titan IV PLF, along with a Cassini spa_simulator and a Centaur simulator. "1_¢ lower lx_t'tion

of the spacecraft simulator was a high fidelity

developmental test model ('DTh0 suppliedby YPLInc.ludcdittldswas one RTI3 dynamic simulatorand

two RTG mass simulators.The upper portionof the

simulator and the large High Cain Antenna

(HGA) at the top of the s[mcccraR were simulators

[xovided by LMA to represent the proper geometry andvolume effects. Figure 7 is a photograph which shows

the Cassini spaog:rafl simulator and the aft section ofthe Titan IV PLF in LMA's acoustic chamber. Figure 8

is a photograph which shows the acousticblanketsmounted on the intm'io¢ isogrid wall of the Titan IV

PLF.

Phase Two testing consisted of a series of sevenacoustic tcslstodetermine the acoustic cnviromncnt and

the RTG vibrationc_ent for three diffca-ent

blanketconfigurations(3" baseline, 5" vto and 6" VS).

F'_,urc 9 shows the test matrix.

In the test matrix, the term "full coverage" does not

imply 100% blanl_ o_erage but is meant to conveythat the tested configuration had similar blanket

covet-age to thax expected for the actual C.assini flight-

Partial ooverage was a test condition equal to 75% of

the full coverage. Two pardal coverage tests were done,one for the baseline and one for a bamer blanket, to

imp¢ove our _ and pcediction of the effeasof blank_ coverage oQ the tK.F's inte¢ior acoustics.The remflls of these tests are outside the scope of this

'me testmatrix also shows testing with and withoutTVAs. TVAs are tuned vibrationabsorberswhich were

attached to the lower portionof thc spacecraRinan

attemptby JPL toreduce thevibrationof theRTGs.

Again theseresultsare outsidethispaper'sscope of

intere_ however JPL's Cassiniprogram officehas

decxied not to utilizethe TVA design for tl_ C.assini

flight.

Figure 7. Cassini Spacecraft Simulator in Aft Section of 23tan IV PLF

(From Reference 2.)

Figure 8. Interior View of Acoustic Blankets Mounted on Titan IV PLF's

Isogrid Wall (From Reference 2.)

Blankets

1 3-in Std

2 3-in Std

3 3-in Std

4 5-in V-10

5 5-in V-10

6 5-in V-10

7

]Rip=re 9.

6-in V-5

Coveraqe

Full

Full

Partial

Full

Full

Partial

Full

P/L Simulator

No

Yes, w/TVAs

Yes, w/TVAs

Yes. w/TVAs

Yes, wloTVAs

[Yes. w/oTVAs

Yes, w/oTVAs

A

PLF Station "--"

Equivalents Zones/701 _/R

641 2"//13-intor=

2"21' i a,, .Cases _

f3-in, V-1and V-5Blanket

370-_ 'OptionsII312

263 11 t

Test Matrix for Full Scale PLF Testing

The Pbmc 1_o test _ was demB_ed to mcasBr¢the dctta effect of the ¢m, Lmc_cms using new barrier

blmdccts when compared to the ¢mrkOmnem using tb¢bascfin¢ blankets. S_¢¢ the rcvcttcrant acoustic field

of the test chamber is different than the travcliag

acoustic wave at a launch pad, it was felt that delta

measurements would be most meaningful as opposed tothe absolute measurements.

3 Ext Micsevery 120"

5527

4928

432--9

To properly quantify this delta effect a number ofmicrophooes were u_ to measure the PLF's interior

acoustic field, as shown in Figure 10. A large numberof these micmphon_ were located in Zones 9 and lOof

the PLF, which was the mB_m of high interest for the

RTGs. Other microphones wcr¢ kx:a_d to measure theacoustic field in other zones of the PLF and to reflect

past and futu_ "['RanW flight locations and past testlocations by JPL. One of the microphon_ is visible in

F'_na'e & A small number of accelerometers weremounted on the simulatot_ to ensure that the simulators

were behaving normany. Although not shown here, JPL

aad MDA also bad a large amount of instrdmentatlon to

measure the vibration response of the Cassixd spacea_

and PLF resIxct_ely.

FULL SCALE PLF TESTING gJF,S-_3[S

The z_xtts from the full scale PLF testing were very

successful Referring again to the test matrix of Figure

9, the key tests were Tests 2, 4and 7. Test2

estabtished the baseline measurements using the TitanIV baseline blanket, whereas Tests 4 aad 7 would allowthe cakndatioa of the delta effect of the new blanket

dcsigas abo_ tbc bas¢l_. (I'c_s 4 aad 5 ate

essenti_y repeals from an acous_c point of view. The

/M2

M4 1130 ®

4 M9.10, t

Ill

]Rlpm_ 1@. Full Scale PLF TestImWffimmt_tim (From Rd'_ 2.)

10

pcesence of the TVAs might affea the spacecra_v_ratien response but does not affect the PLF's interioracoustics. Tests 1 and 2 were used to confirm that the

presence of the spacecraft simulator did not causeanything abnormal to ocoar within the PLF.)

"Ihe acoustic excitation on the external side of the PLF

simulated the Titan IV flight external spedficafion andwas based on the average of six control microphones.The test to test repeatability of this external excitationwas extremely good (range of 0.4 dB over an 7 tests at200 and 250 Hz). However to account for even thesesmall variations all test data was a_usted to representthe level which would be obtained if the acousticexcitation was exactly the Titan IV external

specification.

Figure 11 illustrates the main results of Phase Twotesting. The external specification is the desired PLFexternal specification. Test 2 data shows the average of10 microphones in zones 9 and 10. This represents theaverage PLF interior level in the RTG region when theTitan IV baseline blankets are utilized. Similarly theTest 4 and Test 7 data represent the same miorophone

average when the VI0 and V5 blankets are substitutedfor the baseline blankets in zones 8, 9, 10 and 1I.

From Figure 11 one can see that the new blankets werevery stxxxss_ in reducing the PLF interior acoustics to

levels below those provided by the baseUne blankets.Also whereas this improvement is largest at 200 to 400

I-Iz, it is a positive improvement at all frequencies.

Figure 12 illustrates the delta improvement for the V10and V5 blankets. This figure shows that both the V10and the V5 blankets were suocessf_ in reducing theRTG acoustic environment by 3 dB at 200 and 250 Hz.For the Vl0 blanket this improvement is 3-5 dB at 200Hz and 4.0 dB at 250 Hz. For the V5 blanket the

improvement is 3.2 dB at 200 Hz and 4.6 dB at 250 Hz.

Similar values are reached when the test data isevaluated at the t'95/50 statistical leveis, instead of atthe mean value.

Of course, the ultimate goal was to reduce the RTGvibration response to prevent a v_ration requalificationtest of the RTG. The 8nalys_ of the RTG response is

outside the scope of this paper, but an indication of thevibration reduction achieved is shown in Figure 13.

This figure shows the acceleration PSD (power speetraldensity) response at the base of the RTG dynamicsimu_tor for the baseline blanket and for the (6") V5blanket. One can see substantial improvement,

particularly in the 200 and 2.50 Hz frequencies.

A complete summary of the full scale cylindrical PLFtesting may be found in References 2 and 3.

_s0 1.4

1

"4

,4

'_1 .4[]

i,,°t

Figure !1.PLF Zone 9-10 Mlerephone Averat,e for

Baseline, V$ and ViO BlanketLq

;0s

;00 1 ..... I ..... I

I0 I00 1000 100013

I _ Ext Spec. (150.4 dB) -'--°---" Teng #2 (l_,eliee) ""-°"- Te_ #4 (V I 0) --"*---" Te_ #7 (VS)

11

Figure 12.

Test Meamred SPL Reducllen for PLF Zom_ 9-10,

Utilizing V$ sml VIO Bhmke_,s

5

45 .

4 •

35 '

25 •

2

IS .

1 .

05 '

0

I0

..... I .... I ..... I

100 1000 10000

IoVIOBlaidum(Tmil2-Te_#4) --.e.i. imlpmvementduemVSBla_lr.e_(Test#2.TesZff'/) JI

lop

iO.J

U

"_ lO.Z

JUu !0 "3

1U

10.4o

,i--o.,,--. irnlx_vcm_., m due

iO'S

I0

:.'-:--:--'---I.........J-'|-!-!-_J-l-l'v-",-....._: :'-::I:::I:i::l

..............l........J .........Bueline - No TVAs -11 l-i

.......................I............I -I]I

:::::::::::::::::::::::::::::::::::::::::::: ::: : :!!! i!?!!i:/.......................IZ}.I .........- .................. L!IZLZIZZi.i....i...................... j:.... I

:'j=-_-:._:-- • _: ._!!_!!!-.-:!;!!_:!!_!!.!!_!:!!!!!!!kz:!_!!!!:!_.!;_!! 'r:!!i

:. " 7._:'...............................................i..

..........._ ...._-....---'._...................................

• i " "

........................................_.l..:..........Frequency(Hz)

I00 10g0 100O0

RII_ t3. /udal_gbrltk= ]P..edactlonMmma'ed•,*RII'GIbm _ Rd=z'm¢e 3.)

12

BLANKET SELECTION FOR CASSINI MISSION LESSONS LEARNED

The tectmical assessment of the Phase Two test data is

that both of the new barrier blankets ('/5 and V10)

exceeded the goal of reducing the acoustic environment

by more than 3 dB and significantly reduced the RTG

v_ration response, at the 200 and 250 Hz critical

fxequencies. No detrimental effects were seen at any

frequency or in other PLF zones.

The technical assessment of the test data is also thatboth of the new barrier blankets had _filar acoustic

performances and that other programmaticconsiderations could lead to the selection of the final

blanket design for the Cassini mission.

NASA LeRC's Cassini Project Office has selected the

v5 (6"thickblankets)fortheupcoming Cass_mission. Factorsweighed inthe decision,inadditionto

the acoustic improvement, were the added weight of the

barrier blanket systems, and contamination, separation,thermal,venting,and clearancefactors.With most of

theseconsiderationsbeing nearequal the weightof the

blanketsystem became the decidingfactorand the

"lighter"V5 blanketwas chosen over the heavierVI0

blanket.The %'5blanketisstillapproximatelyfour

times the weight of the Titan IV baseline blanket.

Because of the success of this blanketdevelopmental

test program, vibration rcquafificafion of the RTGs forthe Cassinimissionwillnot be necessary.The

_rion of the new V5 blankets to reduce the acoustic

excitation and the subsequent vibration of the RTGseliminates the need to manufactm'e additional RTG units

for a requalification test program, thus saving

approximately $20-25 million in manufacturing cost and$5 million in testing cost.

The two phase test approach used to solve the Cassini

mission's problem was extremely successful Numerouscandidateswere quickly evaluated in the fiat panel

testing and when they were found to be --_t_factory,additional candidates outrode the original limits were

found tobe promising. These promisingcandidates

were then tested in the full scale PLF testing and found

to exceed the original goals of the blankettestprogram.

If the initial lx'oposed new blankets were not first tested

as flat panel samples but instead tested only in the fullscale test and there found to be unsatisfactory, a large

amount of time, money and effort would have been

wasted in pe_orming the full scale tests on theseblankets.

Knowing that there is a difference between a small flat

panel sample and a flight cylindrical PLF, is it possibleto use the results from a flat panel test to predict theresults in a full scale PLF test?

Analytical software codes, such as PLFNOISE andVAPEPS, can [xedict acoustic levels within the PLF

using the blanketcharacteristicsalong with the PLF

structural and geometric properties. To obtain even

quicker ptedicn'ons during the flat panel testing, a

relatively simple method was developed by CambridgeCollaborative and NASA LeRC. This method enables

one to predict the delta improvement of a new blanket

design over a baseline blanket design for a PLF

configuration using the fiat panel sample test data.

Using dynamic power balance and assuming steady stateconditions and that the energy absorbed by the blanket

is much greater than the energy absorbed by theunblanketed PLF isogrid wall and by the spacecraft,

then the following equation can be derived:

+ lOLogto

÷ l-S*

+ l-S"T _ow blaake¢ --

S"

where,

A = Improvement (dB) of New Blanket above Baseline Blankettg = Measured Blanket Absorption Coefficient

z = Blanket Transmission Coefficient = i0 _--_-6=)

TL = Measured Transmission Loss (dB)A TL = TL of Blanket with PLF Isogrid - TL of PLF Isogrid

s" -- I B2_et_ suzzace_____ea__oz_£LFk Total Surface Area for PLF ]

13

"Ibis equation also assumes that no acoustic energy islest m-ucUa_y through damping or v_rationmechanisms and that the power balance is valid withineach _ band.

The delta improvement is due to two facto_ The firstfactor is the ch_ge due to the new abscrpti_characteristics. The second factor is the chan_e due tothe new trammimion Ires dmwact_ As stated

earlier, in develop/rig the new Cassin_ blanket, we hadto minim_ our decrease in the tim fact_ (keep the

base/ine absoqX_) and maxim/ze our increase in thesecond factor (increase transmis_'on loss).

"[he measuxed data from the fiat panel tests for the TitanIV basetine, V5 and vl0 b_ekets can be used wit), tim

_ltmt_a to ix_lict tl_ imlXOVcmcat_ for V5and V10 blankets. Tais [xe_ctkm can then be

compared with the actual imgovement (SPL re_x:tion)measured in the full scale PLF tests for the V5 and V10blankets in the PLF zones 9 - 10.

F'_Lre 14 shows the lXedicted verses test dataimlxovement over the baseline blanket for the V5blanket. In tl_ case, the ixedicu'on methodology results

in an un_ctim at all frequencies. "[he shape ofthe lxediction spectrmn does follow the actual testspectrum wen, with both the _ and actual testdata peaking at 250 Hz.

A _ comparisoo for the improvement due to theV10 blanket is shown in Figure 15. This comparison isbetter, however now the prediction tends to be a slightovezixedicfion at the f_gluencfies of greatest interest(20o-25o Hz). Aga_ the mape of the predictionspectrum follows the actual test spectrum well, withboth spearums peakingat 250 I-tz.

It is not clearly _tood why the pt'edictionmethodology results in an _ction for V5 and anoverlXedicfion for Vl0. The answer may lie in theinhezent assumlXiam of the methodology. Or it may bebecause the Vl0 blanket rcsulls depend more on thetmnsmi_on loss fa¢_ than the V5 blanket results and

that the flanking paths may have differed slightly in thePhase One and Phase Two test setups.

Nevertheless these com_ showing reasonablemagnitude and frequency corre_on with test data,gives us amfide.nce that this pt-ediction methodologymay be used to give a first order approximation on howa blanket design would perform in a full scale PLF test.

PREDICTING BI.ANKET PERFORMANCE

All predictions given are delta impmv_mmts above the'I_tan IV base,linc blank_ (3 inches tlnc_ w_h nobarrier). That is, the interiora,zuuics (sound pressurelevel, SPL) of the PLF will be reduced when theIxedicted delta imgovement is positive.

Figure 16 shows the effect of increasing the blanketbatting thickness and introducing a ban'icr that iscentered in the blanket Refer to Figure 3 for details of

the blanket designs. The DOE 1 a_rve shows thatadding one mote inch (4 inches) of batting results in asmall improvement. A 0.24 psf barrier has beenintroduced in this 4 inch thick blanket in DOE 4. This

tesu]_ in an imlxovement around 500 Hz, but isactually less than DOE 1 (no barfie0 below 315 Hz._ the ban-letdmsity up to 0.44 psf f_ the 4inch thick blanket further imlxcves the blanketperformance as shown by the V1 prediction. F'mally,for V5, the barrier remains 0.44 psf, but now theblanket thickness is 6 inches, with I inch being addedon each side of the barriex. This helps impt'ove both theabsct'ption and transmission Ices and results insubstant_ improvement.

F'_ure 17 shows the effect of the location of the ban'ier.The V1 curve shows the prediction when the 0.44 psfbazrier is oentetedin the 4 ind3 blanl_L Moving the

0.44 psf barrier one inch toward the PLF's interior,results in the _ given by the DOE 7 curve.Some _ in tram_ loss is more thnn offset

by the reduction in _ for this design fn>m 225to 630 Hz. The curve predicted for when the 0.44barrier is moved one inch outward toward the PLF

isogrid waU is shown in the DOE 8 curve. O"hebattingdmsity has also changed.) At the frequencies of interest(200 - 250 Hz) the wansmission loss factor is nowsignificantly smaller and the increase of the absoqXiondoes not overcome this, resulting in a negative ch,mge,.If the barrier is left in this position but is made heavier

(0.88 ps0 the _ission loss factor improves and theabsorption factor remains the same. Tais is slmwn inthe V12 lxediction curve. (Again, the batting densityhas cringed.)

The flat panel test dam _ in _ Onc testing isvaluable information. The _ methodologyillustrated in this sectkm is one way of using this data toundcmm_ some of the coocepts of blanket design

which were learned during this program.

For all the bl_ket predictions, the fiat panelp_sented in Reference 1 are used. When a blanket was

14

5.00 •

4.00,

350,

3.00'

*_ 2.50

I_ 2.00

I._

I,_

O.SO

100

Figure 14.Predicted versus Test Measured SPL Reduction for PLF Zones 9-10, Utilizing V5 Blankets.

/ _L. "

I I I I I I I I

125 160 2(]0 250 315 400 500 630

Frequency (Hz)

I _ V5 Prediction 1.4..-- V5 Measun_d I

I I

800 1000

Figure 15.Predicted versus Test Measured SPL Reduction for PLF Zones 9-10, Utilizing VI0 Blankets.

5OO

450

400

3,50

_ 3.00

_ 2.50

_ 2.OO .

I Ni.m L

1

0.50 [0.00

100

T

I

I

I

I

\,l\

I I

125 160

I I I I I

200 250 315 400 500

FI't_luenc'y(Hz)

--o-- YlOPRd_ction ----°--"YlOMe_,med Im

I

I I !

_ 1000

15

g

F'qgure 16.Predicted SPL Reduction for Various Blanket Designs

showing the Effects of the Birrier's Presence and Density.

-0.50 .4.

100 125 160 250 315 400 5(}0

Frequency (Hz)

-='0"-- V5_DOEI _r'_" DOE4 "--"¢---- Vl

630 800 10130

4O0

Figure 17.Predicted SPL Reduction for Various Blanket Design

showing the Effects of the Barrier's Location.

160 200 250 315 400 500

Frequency (Hz)

oVI_DOE7 _DOE8 "_°'Vl2

630 10{X)

]6

retested,a straightnumericalaverageof the datafrom

the two runs were used inthe pt'ediaion.A valueof

0.765 was used forS* inallpredictions,which istypical

ofthe Cassiniflightblanketcoverage inthePLF zones

of intcrest.

CONCLUSIONS

A multi-organizational effort, led by NASA LewisResea_ Center, to develop and test verify new acoustic

blankets has been sts3cessfully completed. Two Right

viable blanket candidates, configurations V5 and VI0

have been found which meet the goal of reducing the

PLF's interior acoustics in the zones of interest by 3 dBor more at the Cassini mission critical frequencies of

200 and 250 Hz. The V5 blanketshave been selected

by the Cassiniprogram tobe utilizedforthismission.Because of thissuccess, the Cassini'sRTGs do not have

to be vibration requafified, resullmg in $25 - 30 million

dollars in savings for NASA.

The two phase test program followed in this effort wascritical in meeting the objectives of the test program. In

Phase One, numerous blanket candidates were quickly

evaluated by fiat panel testing to arrive at potentialblanket candidates. In Phase Two, theseselectblanket

candidates were then tested in a full scale cylindrical

payload fairing to determine tbcir performance in a

realisticflightenvironment.

A wealth of acoustictestdatawas obtainedduringthis

testpv3gram. A methodology forusingfiatpaneltest

data to obtain a first order lXediction of the pedofmanceof an acoustic blanket in a PLF has been _.

The information lxesented in this paper may be u_Jzed

for other space missions and their own special

applications which may differ from the Cassinimission's. The barrier blanket tecimology developed for

this program may also have non-space applications forcre_ding quieter acoustic environments for automobiles,

ships, airplanes, homes, offices and indusWial settings.

REFERENCF_

I. Armel, T.L, July 1994, _I'itan IV Payload Fairing

S/N 000036 Acoustics Flat Panel Tests," M_onnett

Douglas Aerospace - Space Systems Report MDC94H0067.

2. Bradford, L, October 1995, "Cassini Payload Fairing

(PLF) Acoustic Bla_et Test, Part B: Test Report,"Locktmed Martin Report IQAS3-00014.

. Bet'gen,T.F.,July28, 1995, "CassiniPartial-

DTM/Titan IV Payload Fairing(PLF) AcousticTest

Report,"JetPropulsionI.zboratoryInteroffice

Memorandmn 325D-95-1ff2.

ACKNOWLEDGEMENTS

The authot_ would like to acknowledge the contributions

of the following people who contributed to the success

of this test program: Pat Symons, Bill Taylor, JamesRobinson and Kuan Lee of the NASA LeRC Launch

Vehicle Project Office who provided the funding and

project guidance for this ambitious test program; IAloBradRx_l, Bob Foster, Abe Jack, Tom Sayuk, TomStates and the entire LMA team who performed the

Phase Two testing and analysis; Theresa Atmel, Bob

Kessler, Mary Long, Mike Seely,George Smuffex and

the entire MDA team who designed the blanketsand

directed the Phase One testing and analysis; Even Ht_t

of Analex _ and Dr. Jerome Manning andBen Hebert of Cambt'idgc Collaborative, Inc. for their

trem_ test support and data analysis throughout the

test program; ThornBergen, Pan1Hardy,HarryHimelblau, Dennis Kern, and the JPL test team for their

insight on the Cassini spacecraft design and their testdata analysis; Don Wong and Norm Lagerquist of the

_ for their Titan IV vchJde

knowledge and support; and John Kopec and PeterStrauss from the Riverbank Aamstical Laboratory for

performing the PhaseOne testing.

17

FormApprovedREPORT DOCUMENTATION PAGE OMBNo.0704-0188

pul_c mpmlin9 IxJclenfor thls ¢oilo¢lkm of InlonmlJon Is mtknmed to rage 1 hourpw mq)onee, Includkl0_e _ne ior mvievdn0instnclJo_, seamh/_g ,--blino dala soumm,9athldno and _lno the dals minded, and ¢xxqlM(J_ and mvwwmg Ihe coitec_n of InfocmalJon.Send ¢_rmmlS _ this bu_en esttmale or any ottw asped of miscoCec_Jonof _. indudinO suggmUms for mdudno mlsburden, m WahlnOton _ 8eMms, Otmcmmm f_x Irdorms_n OpenCom and Repor_ 1215_Davls Highvmy. Suite 1204, Argn_ VA 2220_-43(_. and to 1heOffice of Mana_ 4m¢1Bud_ Pa_ Reductlo_ Pro_ (0704-01U). Wash_, DC 20503.

1. AGENCY USE ONLY (Leave blam*) 2. IEPORT DATE

July 19964. TITLE AND SUBTITLE 5. FUNDING NUMBERS

Cassini/Tilan IV Acoustic Blanket Development and Testing

eL AUTHOR(S)

W'dliam O. Hughes and Anne M. McNelis

7. PERFORMINGORaMgZATK)NNAME(S)ANDADDRESS(ES)

National Aeronautics and Space AdminiswationLewis Research CenterCleveland, Ohio 44135-3191

9. SPONSORINGRBONn'ORINGAGENCYNAME(S)ANDADDRESS(ES)

National Aeronautics and Space Administration

Washington, D.C. 20546-0001

& REPORT TYPE AND DATES COVERED

Technical Memorandum

WU-256-4X)-_

I 8. PERFORllI_ ORGANIZATIONREPORT NUMBER

E-10331

10. SPONSORING44ONITORING

AGENCY REPORT NUMBER

NASATM-107266

11. SUPPLEI_NTAI:W NOTES

Prepared for the 42nd Annual Technical Meeting and Exposition sponsored by the Institute of Environmental Sciences,Orlando, Florida, May 12-16, 1996. Responsible person, W'llliam O. Hughes, organization code 4310, (216) 433--2597.

12L DISTRIBUTION/AVAILABILITY STATEMENT

Unclassified *Unlimited

Subject Categories 71, 18, 15, and 39

publication is available from the NASA Center for AeroSpace Informafi(m, (301) 621-(}390.

12b. DISTRIBUTION CODE

13. ABSTRACT (Maximum 200 movie)

NASA Lewis Research Center recently led a multi-organizational effort to develop and test verify new acoustic blankets.

These blanketssupport NASA's goal in reducing the Tilan IV payload fairing internal acousticenvironment to allowablelevels for the Cassini spacecraft. To accomplish this goal a two phase acoustic test program was utilized. Phase Oneconsisted of testing numerous blanket designs in a flat panel configuration. Phase Two consistedof testingthe most

promising designs out of Phase One in a full scale cylindrical payload fairing. This paper will summarize this highlysuccessful test program by providing the rationale and results for each test phase, the impacts of this testing on the Cassinimission, as well as providing some general information on blanket designs.

14. SUBJECT TERMS

Acoustics;Acousticblankets;Blankets;Cassini,Payloadfairing;Spacecraftacoustic

environment; Sw.ecraft acoustic tesdng; Titan IV; Vibroacoustics

17. SECUmrvCLASmRCXnONOF REPORT

Unclassified

NSN 7540-01-280-5500

18. SECURffY CLASSU=ICA_IONOF THIS PAGE

Unclassified

19. SECURITY CLAS_FICATIONOF ABSTRACT

Unclassified

15. NUMBER OF PAGES

1916. PRICE CODE

A0320. UMITATION OF ABSTRACT

Standard Form 296 (Rev. 2-89)Prescrlmd by ANSI Std. Z39-18298-1o2

_i C_rl ° _

o _C_Z a_

CD