Additional Mine Classification Capabilities for the INSSFinal Group Report under Grant No. N00014-99-1-0171

Min Fong ChangCharles M. Loeffler

Applied Research LaboratoriesThe University of Texas at Austin

P.O. Box 8029Austin, TX 78713-8029

phone: (512) 835-3444 fax: (512) 835-3259email: loefflernaarlut.utexas.edu

min(aarlut.utexas.edu

ADVANCED TECHNOLOGY LABORATORY

10 September 2003

Approved for public release; distribution is unlimited

Prepared for:Office of Naval Research

Arlington, Virginia

20030929 03Z

Form ApprovedREPORT DOCUMENTATION PAGE OMB No. 0704-0188

Public pottng burden for this coliection of Information Is estimated to average 1 hour per response, Including the time for reviewing instructions, searching data sources,gathering and maintaining the data needed and compieting and reviewing the collection of Infomation. Send comments regerdingthis burden estimate or any other aspect of this collectionof Infmation, Inciuding suggestions tor reducing this burden to Washington Headquar=ers Service. Directorate for Information Operations and Reports,1215 f n Davis Highway. Suite 1204, Arlington, VA 22202-4202, and the Office of Management and Budget.Paperwork Reduction Project (070441M) Washington, DC 20503.

PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From. To)10-9-2003 Final Dec. 12, 1998 to Jun. 30, 20034. TITLE AND SUBTITLE 5a. CONTRACT NUMBERAdditional Mind Classification Capabilitites for the INSS

5b. GRANT NUMBER

N00014-99-1-01715c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBERMin Fong ChangCharles M. Loeffler

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATIONApplied Research Laboratories/University of Texas at Austin REPORT NUMBER

P.O. Box 8029Austin, Texas 78713-8029

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR'S ACRONYM(S)Office of Naval Research ONRBallston Centre Tower One800 North Quincy Street 11. SPONSORING/MONITORING

Arlington, VA 22217-5660 AGENCY REPORT NUMBER

12. DISTRIBUTION AVAILABILITY STATEMENT

Approved for Public Release; distribution umlimited

13. SUPPLEMENTARY NOTES

14. ABSTRACTAs a diver scans a shallow water or very shallow water (SWNSW) area with an INSS high frequency sonar, many objects may be detected or imagedIn the scene. The objective of this project Is to develop algorithms that capture the broadband echo responses from these objects detected by theINSS and extract special echo features to assist In target discrimination from the background. The algorithms are based upon geometric acoustics,broadband array and signal processing techniques, and the physics of elastic waves on thin shells. Initially, three algorithms were Investigated foracoustic robustness. These were, Shell Thickness Resonance (STR) Frequency Notch to estimate the targets shell thickness, Local Target toBottom Multi-path Echo to discriminate cylindrical and spherical objects, and Multi-channel Phase Comparison (MPC) to estimate the target's

height. The third algorithm. MPC, was the most acoustically robust, but required a modification to the INSS array geometry. The first algorithm,STR, was sufficiently robust to be implemented within an INSS unit and tested with Navy and ARL:UT divers. The implementation requiredreductions in the algorithm's capabilities to fit within the INSS hardware and software architecture. Three sets of diver test were conducted in LakeTravls Texas and Coronado California. The final recommendation was to not modify the current operational systems but to consider the STR and

MPC algorithms as part of the target sensing and discrimination suites in future Implementations of broadband sonar systems.

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 8. NUMBER 19a. NAME OF RESPONSIBLE PERSON.REPORT ABSTRACT THIS PAGE ABSTRACT OFPAGES Charles M. Loeffler

U U U 63 19b. TELEPHONE NUMBER (Include ara code)512-835-3494

Standard Form 298 (Rev. 8-98)Prescribed by ANSI-Std Z39-18

Additional Mine Classification Capabilities for the INSSExecutive Summary

Overview. This task was part of a set of P31 efforts to increase the capability of the diverhandheld sonar, INSS. This task was an investigation and implementation of a few targetdiscrimination techniques, which had been demonstrated in laboratory tests. The goalswere to assess the robustness of each technique in operational environments and whenimplemented within the constraints of the sonar system's hardware and software. Eachtechnique investigated in this project detects particular high- frequency broadband elasticand/or specular echo response feature, extract specific echo feature parameters, andprovide additional target discrimination clues to the operator. Three echo features wereinitially selected for development:

"* Shell Thickness Resonance (STR) Frequency Notch: This technique measures afrequency dependent target response. It first makes a thin-shelled objectdiscrimination and then estimates the shell thickness. [1,2,3,5].

"* Local Target to Bottom Multi-path Echo: This technique measures the relative timedelays of multiple echo returns fom the target and local multi-paths between thetarget and the bottom. From this set of echoes, it makes a circular object (e.g.cylinder) discrimination and then estimates its radius.

"* Multi-channel Phase Comparison (MCP) (a. k. a. vertical mono-pulse processing [4]):This technique measures the instantaneous vertical phase front of the return echo andfirst identifies phase discontinuities to discriminate object or object parts above thebottom. This can provide an object height estimate, which can often be the mostdistinguishing feature of a mine-like object when discriminating against high bottombackscatter.

Laboratory Algorithm Evaluations: The first set of evaluations used a surrogate singlebeam data collection sonar system was assembled to collect data to develop and evaluatealgorithms based on these target characteristics. The surrogate sonar had an arrayconfiguration that was similar to the handheld sonar. The data were recorded andprocessed offline in these tests. Algorithms were evaluated for their acoustic robustnessand INSS implementation compatibility. Algorithms showing the greatest robustness andfeasibility were then selected for INSS test implementation.

An initial acoustic database was collected for these evaluations in a controlled indoortank environment for algorithm development. Targets were placed on a circularunderwater 8 ft diameter rotating platter. The targets and environment were insonified,and images were collected over a complete 3600 of aspect angles (Fig. 1), which allowedthe evaluation of the algorithm's aspect dependencies. The frequency band of operationwas chosen to match that of the INSS and covered nearly one full octave of bandwidth.Targets tested included reference objects with simple geometry (e.g. cylinders orspheres), natural objects (e.g. rocks of different shapes and sizes), different bottomsurfaces (e.g. smooth sand and rocky), and several mine-like shapes (moored targets,cones, complex shapes).

2

Aluminum 1/4" thick cylinder

Objects on sand Objects on rocks

Figure 1: Data collected at the ARL:UT laboratory platter: These imagesillustrate the typical of the data collected by the surrogate sonar fordeveloping and evaluating the algorithms. First pair shows an easilydistinguished echo on a clean sand bottom, while the second image setshows similar objects on a bright rocky bottom. The rocky bottom hidesthe targets, but they are discriminated with the STR algorithm.

Three techniques were evaluated. Of these, only the 'Local Target to Bottom Multi-pathEcho" algorithm was insufficiently robust to continue into the next phases of testing.Both the 'Shell Thickness Resonance (STR) Frequency Notch" and "'Multi-channel PhaseComparison (MCP)" algorithms demonstrated sufficient acoustic robustness to beconsidered for further evaluations.

3

The STR algorithm exploits distinct high ka features in the spectrum of the back scatteredecho to estimate the material properties and shell thickness. FFT's are used to to estimatethe spectral shape. The fidelity of the esitmates improved with increasing FFT lengths.Algorithms based on shorter length FFT's lose some fidelity and can only estimate theshell thickness for assumed material type. These are primarily "thin-shell detectors."

The MCP algorithm was the most acoustically robust of the three techniques. It estimatesthe vertical angle of the phase front back scattered echo. Height changes of the bottom orobject cause rapid changes in the phase front angle. Jumps in this angle can be used toenhance object detection, particularly in highly reverberatant environments. And, with afew realistic assumptions, the angle can be used to estimate the height of the object. Thealgorithms that compute height estimates are relatively simple and robust. (Versions ofthese algorithm are being used in the most recent class of Navy obstacle avoidancesonars.)

Both the STR and MCP were selected to continue the evaluations based upon theiracoustic robustness results in the laboratory tests with the surrogate array and offlineprocessing.

Algorithm INSS Implementation: Two selected techniques were for the next set ofevaluations. These would be implemented on an INSS operational or ADM system todetermine their ability to operate within the system constraints with minimalmodifications to the operational hardware, software, and user interface. Early on in theevaluations, in discussions at the review meeting with the EOD and ONR programoffices, it was decided to not pursue the MCP 3D technique. This technique wouldrequire an additional array element and signal electronics channel, which was outside thescope of the overall INSS P31 program. Future implementations should consider it sinceit is a very robust technique.

The remaining technique, STR algorithm, was implemented on an INSS ADM system.To install the STR algorithm into the INSS unit, the algorithm was compressed to matchthe INSS front-end architecture and adjusted to operate in real-time. The constraints ofthe processing hardware required that the fidelity of the algorithms be reduced. Forexample, FFT lengths were shortened to reduce the processing load. This increased thefrequency bin sizes and reduced the algorithms ability to detect and measure spectralnotches, which reduced the accuracy of the thin shell discrimination and thicknessestimates. The STR algorithm was successfully implemented into an INSS ADM systemin a real time configuration and tested in the laboratory test facilities where the originalsurrogate sonar data was collected. After successful completion of these tests, the STR-enabled INSS prototype unit moved to a series of diver tests.

Developing an INSS display and user interface for the implemented algorithms was anadditional task in this project. The goal was to provide the diver with the appropriatevisual cuing on the display that would indicate the discrimination algorithm had classifieda target without impacting the basic INSS functionality. A number of prototype displayswere developed on a desktop computer using the laboratory data and discriminationalgorithm's classifications. These prototype displays were presented to Navy divers whoprovided a number of concise, specific, and clear comments that were incorporated into

4

the display implementations on the INSS ADM unit. In the final implementation, one ofthe diver selectable modes initiates the target discrimination processing. When thediscrimination feature is detected, a small color-coded mark is placed on the target on theimage. The divers would typically rescan the target to determine the consistency of themark and then store the resultant image.

Diver Tests: A series of diver tests were conducted to test the STR algorithm'scharacteristics and robustness in multiple environments (Fig 2). In the first set of divertests, ARL:UT and Navy divers tested the STR-enabled INSS prototype unit in acontrolled environment (Fig 3). The divers tested the unit at the ARL:UT outdoorredwood tank facility against reference targets, mine- like shapes, and natural targets(boulders and a limited rocky field) resting on a smooth bottom. In these tests, 4 out ofthe 5 divers found the STR algorithm and display beneficial for mine-like (thinnedshelled) objects versus non-mine objects (rocks) discrimination. They noted that a slowpanning speed enhanced performance and that there was an apparent sensitivity to D/Eangle relative to the target. Navy diver feedback was favorable and provided a guide tothe first major revision.

a) b) C)Figure 2: Navy diver scanning a target at the ARL outdoor tank facility. b)Navy diver receiving dive instruction at the Naval Amphibious Base. c)STR enabled HRLMD unit

5

PDM II Standing cylinder

- 1.-Rock field Diver, PDM, & Rock

Manta Standing cylinder, rock, & rock field

Figure 3: Data collected at the ARL:UT outdoor tank facility: Theseimages illustrate the results typical of the diver tests in the large cylindricalredwood tank with thinned shelled objects marked by the STR algorithm.Very few false calls were made on the tank floor, although the outsidewalls were often marked by the algorithm. The 4 h image illustrates theresults of improper grazing angle.

In the second set of diver tests, ARL:UT divers tested the algorithm in a real-worldenvironment (Fig. 4). The divers tested the revised STR-enabled INSS prototype unit atthe ARL:UT Test Station (LTTS). This is a highly reverberant environment (i.e. rockybottom). Mine-like targets, as well as ideal targets, were resting on a rocky bottom andtested out to a range of 40yds. The algorithm performed well in this environment. TheMK56 mine was consistently marked by the STR algorithm while a bright, (high TS)solid concrete, non- minelike target was not marked. There were very few false alarms onthe limestone and mud bottom with sporadic false alarms on the rock cliff.

6

Cmcretc wedding mc~k & Jarge iteel mine-like "aret C~oncrete wvdding caike & large steel m ine-like target

Figure 4: Data collected at the ARL:UT Lake Travis Test Facility: Theseimages illustrate the results typical of the diver tests on the rocky lakebottom at Lake Travis. Note that the (thinned shelled) mine-like steelcylinder is properly marked, while the bright, yet solid, concrete "weddingcake" is appropriately not marked by the STR algorithm.

For the third and final diver tests, an STR enabled INSS unit was evaluated at the NavalAmphibious Base (Coronado, CA) during the spring of 2002 by Navy divers familiarwith VSM-MCM and NSW missions (Fig. 5). Seven objects; mines, mine-like targets,STR test targets, and a rock; were deployed for the test. Each diver made two dives, ashakedown and evaluation, with the unit. Each completed 2 or 3 scans and saved theimages to cover the object test field. After each dive, the images were uploaded andcomments were logged for evaluation. The eel grass, which was often brighter than thetargets, made it a very difficult environment for the STR algorithm to discriminate theobjects. The STR algorithm consistently marked the STR test targets but not the othermine- like targets (PDM II, Rockan, and buried Manta). The sensitivity of the verticalD/E angle relative to the targets appeared more critical in this test. The divers noted thevertical beam sensitivity, but felt that the STR concept was a "good feasible application"and could help scout swimmers to avoid minefields.

7

Figure5: Data collected in the Coronodo Tests: These images illustratethe results typical of the diver test near the pier at the Naval AmphibiousBase in Coronado, CA. The shallow water and eel grass made this achallenging environment.

Final Recommendation: Two sets of recommendations were derived from the divertests and the overall program. In the short-term, the STR algorithm should not beimplemented within the INSS. The system constraints imposed upon the algorithm dueto processing load (e.g. shortened analysis FFT lengths) and physical constraints (e.g. thefixed vertical D/E angle) resulted in an algorithm that was promising but not yetsufficiently robust to support an operational implementation. In the long term, it isrecommended that an STR implementation is considered in a future upgrade of the INSSor for a similar AUV implementation. An STR algorithm enabled sonar system shouldbe designed with some specific characteristics that do not constrain its capabilities. Inaddition, the MPC height discrimination algorithm that was tested in the laboratory in thisproject, should be considered for future INSS and AUV implementations. It is a robustalgorithm that can provide a significant discrimination (height) clue in mine and objectdetection.

8

The remainder and bulk of this document consists of a detailed final presentation reportthat summarizes the complete project located in the appendix entitled "Additional MineClassification Capabilities for the INSS" It is composed of the following sections:

I. Objective and Project OutlineII. Algorithms Investigated in Initial Data Collection

- Shell Thickness Resonance (STR) Frequency Notch- Local Target to Bottom Multi-path Echo (triple flash)- Multi-channel Phase Comparison (MCP) (a.k.a. vertical mono-pulse)

III.STR Algorithm INSS Implementation- Algorithm Integration- Display STR Marker Integration

IV.INSS CAD Operational ProcedureV. Navy Diver Tests at ARL TankVI. ARL Diver TestsVII. Navy Diver Evaluation at Naval Amphibious Base, Coronodo CAVIII. Summary and Conclusions

REFERENCES

[1] G. Kaduchak, D.H. Hughes, and P.L. Marston, "Enhancement of the backscatteringof high frequency tone bursts by thin spherical shells associated with a backwards wave:observations and ray approximations", J. Acoust. Soc. Am. 96, 3704-3714 (1994)[2] S.G. Kargl and Phil Marston, "Ray synthesis of the form function for backscatteringfrom an elastic spherical shell: Leaky Lamb waves and longitudinal resonances", J.Acoust. Soc. Am. 89, 2545-2558 (1991)[3] S.G. Kargl and Phil Marston, "Observations and modeling of the backscattering ofshort tone bursts from a spherical shell: Lamb wave echoes, glory, and axialreverberations", J. Acoust. Soc. Am. 85, 1014-1028 (1989)[4] T.L. Henderson, "Matched beam theory for unambiguous broadband directionfinding", J. Acoust. Soc. Am. 78, 563-574 (1985)[5] M. Ezzaidi, D. Decultot, A. Moudden, G. Maze, "Measure of the Thickness of aCylindrical Shell with a Focused Beam", 1994 Ultrasonics Symposium, 1173-1176(1994).

9

APPENDIX

"ADDITIONAL MINE CLASSIFICATION CAPABILITIES FOR THE INSS"

w

00

LLOCo C

0~0

zo U)cii

CUE6-

CO)Ci)6

Z0 CUa

W VCU

LU C 4 - Of4

U) Cu z

cua0

CCo

0- 4- m 4

z 4 '- cu

0. LLU 0

00 0 0 -LU ~ c Cu 4-() = Oý(1 -> cu 4-0 F--0-:

W u c mcu 1IL -C CL4u~ .~ 4- Ea...

LL E ~cn u a) 0 C

40- E

_0 . 0 g -n

0~ E -n _0u CU CUC00 t3 = E v (D .2 3 c_ J

L- U) 4 n =a) CU Ma))

v0 0 4D . Cu -0 a4.1~ a - c a) CU~ Cn-a) _5 -2a)

CU - a)~~ 0 OL 5, _C

0 cu 0) Z ~ a 4- )a)CUc E -- > m ~CUE C

_L r g-C l C na u c C u<~ CU 0 C p L- a) C)/C ucu0 >~C 0 0 I~

o) 0 CO'U

or I I U)*I

0 0

C Lf.

0D U)*- U)U)0 (1)

Cl))

z cu !EE r

004- 0

00

000. oU cui,

Oo 0, 0- a)

CLO 0 a

U) a) -cu 0>

Cu:

0 0

0 a)-) (n c) (u

'(0 (D

Cu CU

EE

Oz 0 0< U (AT U) E

CLCWcn E 0

05 0 0-0 0

cu 0) U) -W0 U)

2-) Cu C 0 C-

V U ) C U ) r -

2 .2

0. 0I- (0

* 0 o

z

00

LL(

Jc)

I-

00Z

LLIIU.CL

zL

-0

L)I

<0

U):

0_j0

0l

Ci) ~cn )0 a U))-wo '

LO

0 c 0

o 00

< 0 cu :3 7D

w 0

zi W _)

0j 0 0

40-. C)U(IL04

00

Cl)L

wP

LIU

Co

w

z

CoL

ARL-ilr fT- A.,i

ADDITIONAL MINE CLASSIFICATION CAPABILITIES FOR THE INSS

TRIPLE FLASH DETECTOR-- LOCAL TARGET TO BOTTOM MULTI-PATH ECHO

Cylinders on rocky bottom

Experimental acoustic image

8

Cose up of triple flash

4

CO))

tu

0 Cow

a _z

_o,•

0-

W

I-

0

"CO

Rocky Bottom Wih Riin ph

10

a)0cn

0 C)

z oo 0

E-- 0

L)B

w0

CO 0) 0

0 U)

Ul) cc

_~0)

CL

(Nrw

0m

C)

0

0.

Cl))

Co

0-

LLJ)

0<

U) -0U

>

C.D

la))

ADDITIONAL MINE CLASSIFICATION CAPABILITIES FOR THE INSS

SHELL THICKNESS RESONANCE NOTCH EXAMPLE

Aluminum 1/4" thick cylinder

Aluminum 1/16" thick object

Acoustic Intensity Image of Cylinder

SSpectrum off of front cylinder flash

0

-5

-10

co

"20S-20

-30

-35

-40 5F rL F( HFrequency (kHz)

CO)

0wz0CO)

Aim

E06.

0

0.1.

(I)

E

0u

Thickness Resonance Theory For Different Shell Materials

Shell Thickness Estimated for Different Frequency Notches Spaced 50kHz Apart11 , , , , , , ,

1 0 .. ... ... ..o 0t Steel.. .9 - . . . . . . . . . . . .. i . . .. . . . . . !.. . . . . . . . . . . . . . . . . . . . .i . . . . . . . . . . A l m. . . . .[. . .

.=_ 8 1+ ::Fiberglass

5r A l m. .. . . .S6. . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . : ... . . . . . . . . . .. . . . . . . . .... . . ....Fb ~ g c s : . . . . . . . . . . . .

(00"F 5 - . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .: . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . .:

CA

3 - ..................................... ......... .................. . -

4 . . -.- +3 X.

+ +

200 250 300 350 400 450 500 550Frequency (kHz)

Frequency Notches for Different Thicknesses Spaced 1/16 inch Apart

600 ,

5500 Steel5 0 0 ....... ............ ................................... A lu m

_ :... ... .... .... . . . .. .......

4 0 0 - .... ..... .. .. ....... ............450 0 + - Fiberglass

450

. . .NQ

2C 50 +........... .. . ..

1+X50

1 0 0 - .. .. .. .. : .. . . . . . . .. . . . . . . .

2 5 ................................ ........................÷ ....... .... ........... ................... .

"2 3 4 5 6 7 8 9Shell Thickness (1/16 inch)

18

C.o CD

0 0 04. a

_ 0

> >

4-'4

Iz 2 C 0

LL Z0 2 CD00 EU, i

01) 4' a)

0 m

a) 0 c

0 L

U) 0 ) 0 cE c -

z () a)CoU) ) E, Eo

E 0)M a . c n a' UU) = C0 3: - -a 6 0

U) -U

Z E -cE L 0a)C CDm -

0.. 0 c~c I* u I~L0a)t U) M/>_ a) ))Z~~ X ) d -

m 2 E 0 w

Z~~~ Z .. E0>5U 0 :t:! >

4) C: eCa_ 0

*0 0u

CO)z

wp

w

0z

LL C

UC)

(U)

wz

z0

L11

cu

cni

0.

U) 0) 0

0 0-c

U) 0 ()U) Z- ~U) )(1

z CL 0 c0 2

U) CD ()%U (D

-) u 0).J~ o

cI-0L.

cn o

2u cZ U))

Cil r 0U cu~

.2 .2 (O

U) 0a0) o, -- oE

U) 0~U.--~ ~ -~0

2 *n CL 0 E~-) 0) 0 E U n~

. 2 a-

CIC Cn (

____________________

CC

~< >

H 0 0L H DHV

00

__ 00

> A? U)4)l

z C4

U)

z

00

UC)

w~Cn

CU r 000 C0,n

r CL~ F~ (Dco --t m 4

c~c

0~0

cr 0

U..Z

-C-

zw C)I= o

00DD

clL) L-05

IE-

C.)~

Zcr

z0

pU)Eoc

LL

CL

z

0

ILN

(0" L- t(0 Co

-zO-w U) 0

oE C0 ov,

z0Oz

C LM

0- LL (0 --- U

4c 0 C

cma

Zw

0*

~LL

coWCO)z

0-

LL

z u

- IxLLw

U) z0.

-Jz0

_j0

0I- C 0

0

1.M 0)0 Ca

t cu cu -

0 4-

0 4- '

U) a) 00 U

75 0 aZ. 0_ >%-.

4) 2)0_0 -W C n-

a) -cCD L- -

Cu 0 0 c0t 4-

0.0 a) cu0. >/) -l c

a)) U)

CL CU C)FL - L U) 0Cla)0 -u

0 0n-~~cu 0 a)Cl) CL -a a) .4-

C)0 c. V- " U) CCl) CL. u C 0) a 4~4- Mu cu a.)

a ) a) a)

0I a) Cu-c 'a c ) .0) CU E

0 c

ou U-) L U

U -CY) LO (0

0

0)

Cuu

L- ) z > %

cu Cu o CL.m-

0. >

co U) 0 4a

um. 0 ~ L.

4-' u)

U) U) ) 0l

oo 9) >.

Cu E4) 44-,.

LM 4)0 'C

Uo -n CuLU

o m 00

0 0 0j 0) 0 CCua)Ccu

0

E~. coVU) C

00) ) U) 4-->) a)

L) 0u n 0

a. Cu

Z0 0 u)-a -E C

a)mCu Cu)

co

Lo0H C

CL

0

%knoca

U)4.0 .0 .•• •ii • •, ..........

"0. m mi¸ ~k•

C,,>•. ....

! C o•...

11

E

U U

CL,E)

0)EL) CD Cl) on o

CL5 2 *- CN

U))

Ec

41r

COC')z

LL.Cl)w

CL

z0

z

o0

0a) cuicCL 000 o

Uj Lf cu U

00 0'C:

0 E "

C- 'DC'

>000H

CL 0 C-4 .1) o

Cl)Cl H cu_ C U

c0 E 0 CU0 (D ) 00

0U) (D ~04U L-L.. CE

L. 2U U - j U)(D U)

Vb. cm ~ ~4- L

H LO a . *LO

Redwod Tak Taget iel

Manta @

inIn

33

00

I-o

TrA0

IIN-

wu

0ý

ce,

Ul)

i-z

.... ....

mi36

wL0z

zzF-

0-Jw

C-,0

0z

00

c

CO i.. ~a)

(0 (3u

zE

COOU

I--

<Co M~ a0ýE

0(/) --CU -4 cc

co cc

z 0.

CU - ) Co

E -(D

a)-) (j

cc

(n )

>0u

Z- 'a

I700

~c.co

0 (D

4- C,

q0 U0 0

0 -0

(Uo a)a)

0o 0 rc

0: E4 00-

cts cc 4-, "

U) >j CU a)aV0

lzz 4-' > CU a

W ;) _D c

0l CU L - C 0c U)

< z cct n 0c-F c1., 3 0 C: CU

L0L . ~ *CU ~ cU CU ~ 0

a) - %I.-

I- c c ccýa)0aU) cc~ "a CU a)C mCD C

00 -W a)) c)C: Un U)C

0o- M0 ) n (D

E~i o i 0 I

0. a. a) 4-

U)e c~)o , a)

o 0 0 c o c 0 - C U 0

0 1.. _ E 7 0U(0 U)Va) CUa)

4.- > C

co1 CD -0 ~'t Q3 0 > a0t ý L oa ) > 0~

4-' in * 0

CO 0 0 CUE CU )a

CCoe

Cl)

w

I-

Co)

<U,

Cz t

z

0

Kn

Cu C N

0 cu

o 0

o) 0 )

E Cc

U) _j

C: < Uwi cu 'a

I-Z L.. .

OL Cu u4

LJo CD

M~ 0)

0) 0

0 10

co 0 u 8

3: U).0 cuv

z ( cooý

2 U) co CU )z cnw Cu CU-

Cu 42 0 c.z 0

Cl)0 coLCu

0 cu )-

0C _

0 m. 75Cl

C: E)

0C.)u

I~CY)

4-

LU >.

I 4- E

c,)o0 (Va,LL - ~ ,C:

U) CL () 0-UuW ) > U)

I- EC) a) 44 U

<c) a) CD (-I- = c

3: acu EZ ~ 0oz U) CUa0

(D CU C.) 0)

C U (V a)

flL VU)U)E0 ) CU C/)

Z) U)

0 CU

0 0 0 t•

CO)zw

I-2

oU.w

CL

(Cl1CoZ)

I 0n

04Q

COu

CL -) a)W4n

0~ 0

COO

U

cEnE0_

=)0 :3 _0>

O~ -ý cU) C:CDa)CD c U) )0i

0n >1 0 (Q)) _

C U

W CL)

in Eli

CL (U

0Ž0 0

0.0

CQ >0 _

uI B- 4- a)

0(W 00 cmECl) <' (n).o

coo 1 Q oC) co a)

z ~ ) flO0 (D-

0 ~ E cu o wE*-cnu E~* 0o 2

a)4- a)00

a)u-0 Co) a)CU )Q o. .1-.

a) a)a C a)j C: cu0"> C >i > c

WZWE0 >> 0 .2 c

a) D D D DD) a

U)

o I-

*04

CoCME 0U

CL C) aSr-c -a C) () 0a,)

ICT 0

cIr)a CUC c2IL ) L ) ~U) C:a, a

4n 1.o 0~~~~~ a,- a ~ .

z in a) EEa0a o a- o~ c m

> >0- > >CU0 0 02(D L:

LIIDLI DD0D) CU

> W~

LLCN CoD

z> 00

cicu0 (

uj 0

0E ci)C0o3WcN

-i D

CCl)

C/)Z

a ) EU

woaŽ) -C

N -- 6.

U) E 0 0U-

z' CU0u E

o 0ECU(L ca

ao ) a)?.' c: 0 co

z ~U) :

0 0 " L C (a) -cD E.

L)z >0 U)

ui E

<C,)C l) c E-E

L 1-0 cu _ 0. 0e

Co N / .ý 0)L)

z 0 )y ) C ,n

00) a)* co E

C/) 0)A 0

0 >C Co =

0~ M>OQc o

E 4' iK 1

z

4))

U) CD

(0~ >-

0 -) 0 L- D

Cl) 0 cu)

wL A) > 00 (

c:0>

OCz E cuc a)~

00c0 (n 4- -0

0 0 C

CL 0 ) CU0) U(>D oo

< U .~E Z

z cn W .cnr L- mV '

.0 .- -Yo 0 _n s ) 4-0 cU)

E. U) 0) EID 'U 5 ) _-

->:Co E~ m0 wU 00

>- m U

>- CO -

(U 0-- ~- U) -(0CU 03 =3 -0

> U0 co LU-~-C) L) U)

o 0 Va- U) CE CL~c . U) 0

0oE a)z 3c

-* 0 0, 2

0i 0

m co muC 0)U) m E>. E

Su C-C -)

U)C Cu c E EuI-~U au C . ) C oUu- 23 _7c4 nC

E - C c~Cu-. 0 a) ) ~ ~2

U) a) a). a) 0Ea) Cu9 0 .C: 0>l Cu 0) E.6"~U

m 0 0 -C. "lWD -u- - 7a5 Cu C U D.r~ U)W 64l)0

rV Cu Cu CO E -uJcn

Cum. ~c: '. > o0 TO 0 CuD f

C- . .c) C >VCu 4- >Mn C.) C C "Eal03 --

(D U) 3) LmC C 0 . c cu E CDU) U) C V a)LE. 0. '0

U) C LV

u-0 ~~ Cu 0 0 0

4- CDucE C u 0 Z C ) 0 U)* Cu Cu Cu0 0) CD C ).± C

co 0 .0 0 o-uO g0w uD E 0 -o 0) ~0~~c 4. C4- L C-) U C u )'Cu .6 C) 0) CEu

CuE C 0 Cu 0 ) 0- Cucn CD- c U) .0

r %UV "l C-V C CDVU V 0 Cu o- LC Cu U) Z )---'-~~. Cu ca C=C u- u0 CuO

Cu At- m CuE u ~ -~C CZ l 0~ _ U) 0) V Cu E 020 .Cu C)< U. ) m' V 0UE0Cu C Cu C0 ( a) Cu :0- C C -(D0D (D -0 0L -0

- ) 0 c C u0: E-.a).m-C - 0 E D

~ I -0U -0~l CC '-Cu

-0 CD

-c a) 0 (

a) 0 0

0 (N

(u

Ca) E -

(/) I-

0

_ E0I- 0 Ez..ctc(n (u

a )

LL. zC 0 -Co

co (n

U)- 0) c

C6 0 CD _ - 3:

04- 0

E :3 E0 cn Eu cZ > 4- 0~ 40

0.

L-U. 0 _ Eo ~co U)0L-L 00 cu< ~_cu C : _0

-c ~ 0CO E C a) 4--

-a- -C. o)

Cl) U) cn ý a) (CU 0)0 -

Zn En CL E .r

a) -r -(U

0U m 0~ EuU:a_ E IUZ~ E -

u I- --

0 a) 0

E 00 6 o>3

DISTRIBUTION LIST FORFINAL GROUP REPORT UNDER GRANT NO. N00014-99-1-0171

Copy No.

I Office of Naval ResearchBallston Centre Tower One800 North Quincy StreetArlington, VA 22217-5660

2 Defense Technical Information Center8725 John J Kingman Road Ste 0944Fort Belvoir, VA 22060-6218

3 Naval Research LaboratoryAttn: Code 52274555 Overlook Avenue SWWashington, DC 20375-5320

4 Jim Ryan(PMS-EOD-3)Naval EOD Technology DivisionCommanding Officer2008 Stump Neck RoadBuilding 2144Indianhead, MD 20640-5070

5 Robert Simmons(PMS-EOD-3)Naval EOD Technology DivisionCommanding Officer2008 Stump Neck RoadBuilding 2144Indianhead, MD 20640-5070