,Aý%D-A275 793

AJ~my RESEARCH LABORATORY

Characterization of a DiffractionGrating as a Simulant of a

Selective Frequency Antenna forRadiometric Applications

by Thomas J. Pizzillo

ARL-TR-303 January 1994

* FEB 14.199 4

E

DTIC QUT INSPEOTM•

;aW 94-04925

2M ubl L e; distribution unlimi94 ex0 •

TKe frAnings in this report are not to be construed as an official Department ofi;:y position unless so designated by other authorized documents.

ct•z•tion of manufacturer's or trade names does not constitute an official•,owrment or approval of the use thereof.

Ge-iffoy this report ;hcn it is no longer needed. Do not return it to the

REPORT DOCUMENTATION PAGE _O Apm.0P I Iw I~. I IwO No. q I"04-O .

1. A2111M UNO 4~110Y slwil• LIioNre W1111 &L lWPONTTrIMAWIDAtIWC€Výt

January 1994 Final, from Oct 1991 to Oct 1992

Characterization of a Diffraction Grating as a Simulant of a Selective DA PR: 61102.AH43Frequency Antenna for Radiometric Applications PE: 611102.1H43

PE: 611102.H4

Thomas J. Pizzillo

,. psss olm imnOaM. AIIm iM*l NO AammpllU a. raoim OAS'"UA I

U.S. Army Research Laboratory M,Attn: AMSRL-SS-SH ARL-TR-3032800 Powder Mill RoadAdelphi, MD 20783-1197

U.S. Army Research Laboratory2800 Powder Mill RoadAdelphi, MD 20783-1197

11. SWRUNWlOM

ARL PR: 611102

12.. DSlminowICAMU1SAvltn IrrllinT Flab. Bmsiunl~ conu

Approved for public release; distribution unlimited.

IL ,AUS7rT tmum 2A~ua

This report details the study of a blazed diffraction grating as a simulant of a selective frequencyantenna for radiometric applications. Specifically, the ability to direct and sweep a beam of millimeter-wave (MMW) energy is explored as a potential means of munitions guidance. A grating was designed andbuilt to direct incident energy into the negative first-order peak of the associated diffraction spectrum. Achange of 1 GHz in the incident energy frequency is shown to produce a beam displacement of 0.3°, apolarization dependence is shown to exist for energy in orders other than zero, and a limited scaling ofthe optical theory of diffraction to MMW energies is demonstrated to be feasible.

14. 11UNICTiM Is mu•m or PAGES20

Millimeter wave, diffraction grating, antenna, radiometer t*P 0ow

17. saminCU~rInoim IL SOCUIY CLASUR1FCAnTu I&. 5EUMA CLASSOIATIOW 2L UWTATM OF NTRCOF OFE OTMS PAGEI OF AE1lUATIL

Unclassified Unclassified Unclassified UL

us-IN

Contents

1. Introduction .......................................................................................................................... 52. Bak ground ..................................................................................................................... 63. Instmentation ............................................................................................................... 94. e entalProcedure ............................................................................................ 125. Results ................................................................................................................... 136. Concl io ......................................................................................................................... 18Acknowledgmen .s... ...... . .................................. 18Distribution ................................................ 19

Figures

1. Spectrum due to a diffraction grating ....................................................................... 72. Schematic representation of interference and diffraction dependence on

grate parameters ................................................................................................................ 83. Spectrum due to a blazed grating ............................................................................... 84. Instrumentation diagram ............................................................................................ 95. Geometry of grate used for reported data ............................................................... 106. Measurement geometry ............................................................................................ 127. Calibration measurement of flat aluminum plate ................................................. 138. 0 to 90° measurement of grate ................................................................................. 149. m =-lof figure 8 ............................................................................................................. 15

10. m = 0 offigure8 ............................................................................................................... 1511. m =- Iof figure 8 ............................................................................................................... 1512. 15* polarization measurement .................................................................................. 1613. 300 polarization measurement .................................................................................. 1614. 45* polarization measurement .................................................................................. 17

Tables

1. Calibration plot parameters ......................................................................................... 132. Figure 8 peak parameters .............................................................................................. 163. Peak power polarization dependence for angles 150, 30°, and 450 .............................. 17

3

1. IntroductionThe U.S. Army is investigating millimeter-wave (MMW) radiationas a means of guiding munitions. A crucial aspect of a guidance sys-tem is its ability to detect and track a target. This requires a radar, fortransmitting/receiving the energy, and an antenna, mounted onhigh-precision gimbals, for steering the radiation in space. Alterna-tively, detection can be done by a radiometer, which discriminatesthe cold space temperature reflected off a target from the much hot-ter terrestrial background. The use of a frequency-selective antennaallows for passive scanning. This system would discern where in itsfield of view (FOV) a target is by narrowpass filtering the receivedbroadband signal. Many narrowpass filters could be used to gener-ate a line scan image of the FOV for target identification. A simplersystem, consisting of only two filters, could be used in applicationswhere complete images are not required, such as final in-flight tra-jectory corrections for tube-launched munitions. If an MMW beam isused to steer them in this manner, future smart munitions could besmaller and less susceptible to environmental factors such as launchdynamics. This report details the investigation of a diffraction grat-ing as a simulant for a frequency-selective antenna.

Accesion For

NTIS CRAMI •

DTIC TABUnannounced 0-Justification..... ..... .

By.. .........Distr-ibution I

Availability Codes

Avail and/IorDist Special

SQUALITY ,33

S , C7,5

2. BackgroundThe principle behind diffraction gratings is well documented anddescribed in any of a number of optics books.1 The specific phenom-enon exploited is diffraction of electromagnetic radiation by a grat-ing of reflecting facets. The spectrum produced by a diffraction grat-ing is created by two distinct phenomena: interference anddiffraction. It can be calculated from the general grating equation,

l(sin .- sin 8j)=m, (1)

where

0. = angle of mth order peak relative to the surface normal,6k = incident angle of transmitted radiation,1 = facet separation,

m = mth order interference peak, andA = wavelength of incident radiation.

Figure 1 demonstrates how the combination of an interference pat-tern (top of figure) with a diffraction pattern (middle) produces theobserved spectrum of a diffraction grating (bottom). The angulardistribution of the orders is determined by the interference of radia-tion from multiple facets. It is dependent on both frequency andfacet spacing. Hence, equation (1) may be used to determine the lo-cation of the interference orders for a given angle of incidence. Theinterference is due to specular reflection from the surface as a wholeand is constrained by

4= 0, (2)

where 0o is the angle of reflection. Diffraction from a single facet isalso a specular reflection and determines the location of the diffrac-tion peak.

The combined spectrum is of little use with a standard grating be-cause of the spreading, by diffraction, of the energy into multiple in-terference orders or beams. For purposes of steering radiation, a par-ticular type of reflection grating, a blazed grating, is required.Blazing refers to the angling of the reflecting facets of the grate rela-tive to the surface normal. This angle is referred to as the blazeangle, 6 b. Blazing eliminates the unwanted spreading over multipleorders by directing most of the incident energy into only one of theinterference maxima. The blazed spectrum can be accounted for bytwo effects: interference of radiation from multiple facets and dif-

1For eamtpie, Mies V. Klein and Thomas E. FurtaA Optics, Wikey, New York (1986).

6

SFrn 1. Spectumidue toa diffraction I I I ( f I (

th_ _ _ _ _ _ _function multipdHe I ..... ....with diffraction 0function produces __________ i_________observedcombitedfunction. (Figre DfIacoadapted from Klein I____________and Furtak.) 0

0C

0

fraction of radiation from a single facet. Figure 2 gives a schematicrepresentation of these two effects. The location of the diffractionpeak is measured relative to the normal of a single facet and is indi-cated as 00, (the minus indicates that the angles are on the same sideof the surface normal):

4- 0o. =2 (3)

Hence, varying the frequency of radiation incident on a blazed dif-fraction grating allows a single MMW beam to be directed andswept in a particular region of space. Figure 3 shows the blazedspectrum as a function of the dimensionless parameter Ab/ I where A1is the "tuned" wavelength of the grate. The tuned wavelength is de-fined as the wavelength corresponding to m = 1 in equation (1).Varying A, the incident wavelength, moves the diffraction patternrelative to the interference pattern, enhancing one or another of theinterference maxima. To determine the feasibility of usingfrequency-selective surfaces as a method for steering radiation, ARLdesigned an experiment to measure this effect.

7

Figure . Scheematic Incident radiationS1Otf Interfernce

diffg- ---- *, normaldependence on gate 80

Othorderbean

srffracdon

O,•mTon

rdiIncident radation F

dEIB~~~mStscon reflrec 7tionT

Surfece. . ... .&

Figure 3. Spectrm_____________________

due to a blazed

(tp) spectrum for :)]m a1, Le., Incident 04 2

radiation equals T Up

taned frequency ofgrate;(bottom) spectrumfor 2 ' 4AA

M = 2v L~e., incident __________________raiTh to twc 4 b 2 O

(center) radiation AA1 jbetween these two.

m ,zeroes In i0 A U 2Ab 4

diffraction pattern T 7=7 T 7cancel all other peaksin interferencepattern. (Figureadapted from Kleinand Furtak.)

3. InstrumentationThe radar instrumentation, the data acquisition software, and the ra-dar control software were developed and fabricated by ARL. Theinstrumentation, diagrammed in figure 4, consisted of a bistaticfrequency-modulated (FM), continuous wave (cw) 93-GHz trans-missometer, a diffraction grating, and Hewlett-Packard (HP) dataacquisition and signal synthesis equipment. The transmitter con-sisted of a 91-GHz phase-locked oscillator (PLO) locked to a 100-MHz PLO. A variable IF signal, 1.5 to 2.5 GHz, at -10 dBm was gen-erated by the HP network analyzer, amplified, and then mixed in asingle-sideband up-converter with the output from the 91-GHzPLO. The mixed signal passed through a filter and then was trans-mitted towards the grating via a 5-cm circular antenna at a power of0 dBm. The receiver intercepted the radiation reflected from thegrate using a 15-cm circular antenna feeding a harmonic mixer. Themixer down-converted the received signal using the ninth harmonicof the source generated by an HP frequency synthesizer. The synthe-sizer and source were kept coherent by the oscillator being locked toa 10-MHz reference signal from the synthesizer. A low-noise pream-plifier was used on the harmonic mixer to set the system noise figurebefore processing by the HP network analyzer. The difference signal

Figure& - - - - - - - -]L'•4LHP synaml~zw Recive

instrumeutation 1I111 H etodI mmo

- - - - - - -u - -IHP dga

I-

U 0 HPIB Interface IhWdWar9and

9~fVt

generated by the network analyzer could not be coherently inte-grated directly. Because of the limited resolution of the synthesizer,the phase of the down-converted signal continuously drifted intime. Therefore, a fast Fourier transform (FFI) was performed on thedifference signal, and the maximum amplitude was recorded. Thispeak corresponds to the beat frequency between the transmitted sig-nal and the resolution-limited harmonic, specifically, 8 Hz. An HPcomputer system was used to control the radar and perform dataacquisition.

A number of grates were studied, each with its own specific geom-etry. The data presented in this report were obtained with the grategeometry in figure 5. The overall dimensions are 30 by 30 cm. Theblaze angle, Ob, is 14.5,. Facet spacing, 1, is 4.84 mm, allowing for 61facets to be machined across the face of the aluminum.

The following guidelines are proffered for creating MMW blazeddiffraction gratings. Given a fixed Oi and taking the derivative ofequation (1) with respect to X and 0., we define the angulardispersion,

d0m md I cos ,' (4)

where dO. corresponds to the amount of shift required and dX is thecorresponding change in wavelength. The cosine argument is nowthe location of the receiver relative to the grate normal as a functionof the integer value m and facet separation 1. The facet separationshould be greater than a single wavelength: two or three timesgreater is preferable. The order is then chosen so that the angulardispersion requirement is met; the higher the value of m, the further

Figure S. Geometry of 30 cmrate used for

reported data.

I 30 cm

10

apart different frequency peaks will be. However, the measurementbecomes more difficult, because the angle between transmitter andreceiver is also increased until shadowing effects are significant.Shadowing effects occur when the top of one facet blocks a signifi-cant portion of a neighboring facet, and edge diffraction begins todominate the spectrum. Once a value of m is determined, and henceOm,, equation (3) may be used to determine 66 blaze angle, which isused in the manufacture of the grate. It is recommended when thegrate is made that mill tolerances be kept tight to avoid surfaceanomalies. I found that noise increased significantly when a lessthan precise mill was used in making some of the initial gratings.

11

4. Experimental ProcedureFigure 6 shows the measurement geometry. The grate was mountedso that rotations could be made in two dimensions: rotation in theplane defined by the grate normal and the bore of the transmit ar,-tenna, as shown in figure 6a, and rotation in a plane defined by thereflecting facets and the incident electric field polarization, as shownin figurc 6b. Figure 6a is the plane for measurements of fieldstrength as a function of radiation angle of incidence, and figure 6bis the plane for polarization measurements. The transmit antennawas set 61 cm from the grate illuminating the center region. The re-ceive antenna was positioned 15.25 m from the grate with its boreperpendicular to the transmitter's bore. All three components wereoriented in a common plane.

Measurements were made with the polarization angle (defined infig. 6b) fixed, and the angle of incidence (defined in fig. 6a) varied.Readings could be acquired in steps of 0.10 with the normal of thegrate surface passing from 0 to 90* as defined by a. At each angle in-crement, the frequency was stepped from 1.5 to 2.5 GHz in steps of100 MHz.

Figure 6. (a TasmitMeasurementGrtZgeometry. (a) aAindicates angle ofincident radiation,and (b) f indicatesangle of incidentPolarization./

Graar Reaeion Reflecting facet

12

5. ResultsTo determine the angle between transmitter and receiver, I first cov-ered the diffraction surface with a flat aluminum plate. The platewas rotated from 0 to 90", indicated in figure 6a, with a measure-ment made every 0.5". The receive antenna was then considered tobe positioned at twice the peak abscissa value relative to the trans-mitter. This measurement, after the mixer response is subtracted,was used as the calibration for determining the amount of powerand location of subsequent peak measurements. Figure 7 shows thecalibration measurement for the data to be presented. The ordinateis a measure of the ratio of the analyzer's received IF to the ana-lyzer's generated IF, and hence is in decibels. The noise level is con-sidered to be -75 dB. The peak at 41.5° is a specular reflection indi-cating that the transmit and receive antennas are separated by 83°,i.e., 7° from perpendicular. Table 1 conveys the pertinent informa-tion for the two plots.

Figure 7. Calibration 0omeasurement of flataluminum plate. -20 92 GHz

---- - ------------93 GHzM -40-

0. -60

-80 . "'

-100 .0 20 40 60 80

Incident Angle

Table 1. Calibration Peak Peakplot parameters. FrquenCY pw 64 center

(GHz) (dB) (0) (0)92.5 -29.18 2.0 41.5093.5 -30.66 2.0 41.50

13

Figure 8 is a 0 to 900 measurement of the grate with a resolution of0.5. Figures 9, 10, and 11 are measurements of the individual peakswith a resolution of 0.1". Peak information is presented in table 2.The data show a shift of 0.3 in the OU beamwidth of the first peak(the half power point) and a 0.7* shift in the last peak for a 1-GHzchange in frequency. The center peak does not appear to shift.

The order the energy has been directed into is known because weknow the separation of the two antennas. Specifically, for each 6i,which is taken to be the center of the 03r beamwidth, the corre-sponding 0o must be on the opposite side of the surface normal. Ex-cept for the center peak of figure 8, this is the m = ±1 order. Hence,the first and last peaks are images. The wider width and lower am-plitude of the last peak are due to the changing aperture as the grateis rotated. The center peak is troublesome, because it does not seemto fit with the theory. Figures 12, 13, and 14 are repeats of the meas-urement of figure 8, except that polarization has been changed insteps of 15. (i.e., 15*, 30°, and 45° polarization measurements, re-spectively). Peak parameters are listed in table 3. These measure-ments show the polarization dependence of the directed energy. Thecenter peak shows little variation with polarization, indicating that itis a flat plate specular reflection corresponding to m =0.

Flgure 8. 0 to 90g 0

grate Resolution is 92 GHz0.5. Peak infor- -20-mation is listed in 93 GHztable 2.

CID -40 m=-1 m=0 m=1

-60

-80 -

-10o

0 20 40 60 80Incident Angle

14

Pisan 9. m -lof0& 92Gis

detectable but not 9 Hdistinct. -2093H

w0 4

L 6

0 1)

10 12 14 16 18Incident Angle

Figure 10. m=0of 0.figure & There is no _ ____92 GHzdiscernible shift&2

-20

04

080

36 38 40 42 44Incident Angle

Figure 21. m Iof 0.figure 8. Note thatshift is in opposite 9 ~direction and is zmore209Gzdistinct than that of ------ 93 GHzm -1 peak. 4

060

60 65 70 75Incident Angle

15

Table 2. FigureS8 peak Peakparameters. Frequency ,th Power AM Center

(Gil) peak (dM) (') (0)9 -1 -31.96 1.5 14.2092.5 0 -31.66 2.4 413592.5 1 -46.97 43 6925

93.5 -1 -3372 1.2 14.5093.5 0 -34.10 23 413593.5 1 -48.48 3.9 68.55

Figure 12 lS- 0'Polmizaflon 9 ~

m ~92 GHz-20

- 93 GHz

O -40-

0 - so

-80

-100r

0 20 40 60 80Incident Angle

Figure 13. SO0polarizationmeasurement 92 GHz

-20 ------------- 93 GHz

• --40-

.- 60

-1000 20 40 60 80

Incident Angle

16

Figur 14.45 0poaizatonmea rm e-20- 92 GHZ

- 93 GHz

-40

"0

-80

-100 • • • , . .

0 20 40 60 80Incident Angle

Table 3. Peak power Frequency mth Polarization angle (dB)polarizationdependence for (GHz) peak 15 30 45angles 13-, 30', and 92.5 -1 -56.36 NA NA45. 92.5 0 -31.30 -32.07 -33.83

92.5 1 -49.82 -55.59 NA93.5 -1 -59.96 NA NA93.5 0 -34.12 -433 -5.6293.5 1 -5131 -52.62 NA

17

6. ConclusionsThe results of this experiment indicate that the optical theory of dif-fraction and blazed diffraction gratings may be scaled to MMW ra-diation, with some limitations. It has been shown that the shift of en-ergy associated with blazed gratings is detectable. For the gratepresented in this study, a OX shift in the negative first-order beamwas realized for a 1-GHz change in transmitted frequency. By choos-ing an appropriate geometry for the transmit-antenna/grating/receive-antenna system, along with the corresponding grate param-eters (i.e., the blaze angle and the facet separation), one should beable to devise a beam sweep useful for guidance applications. Also,when the bandwidth of the transmitted frequency is increased, thereis a concomitant increase in beam sweep. The presence of the m = 0peak in the grating measurements is associated with the specular re-flection from a flat plate. This indicates that the scalar theory of dif-fraction is inadequate when the facet separation is on the order of awavelength (I = I.5, X in this case). This anomaly need not be a hin-drance for the application of this concept to guidance, in that the ge-ometry of the system should be specified so that 0., 0 o, 45o.

Follow-on work to this study should include a configuration inwhich the incident angle is held constant while the position of thereceiver is varied. This would allow a more thorough examination ofthe spectrum. Also, measurements out of the plane defined by thegrate and antenna bore should be considered; these would allowfurther investigation of the polarization dependence that was shownto exist for energy in orders other than m = 0.

AcknowledgmentsI would like to thank Timothy Burcham of ARL, without whose helpthis project would not have been completed. Also, Suzanne Strattonand H. Bruce Wallace are thanked for many thoughtful discussionsand their editorial assistance.

18

Distribution

AU.S. Army Aviation Research & TedmologyDefense Tech Information Ctr ActivityAtt: DTIC-DDA (2 copies) Attn: SAVRT-RCameron Station, Bldg 5 Moffett Field, CA 94035-1099Alexandria, VA 22304-6145 Commande

Institute for Defense Analyses U.S. Army Materiel CommandAttn: J. Ralston Attn: AMCDRA-ST1801 N Beauregard St 5001 Eisenhower AvenueAlexandria, VA 22311 Alexandria, VA 22333-0001

Director DirectorBenet Weapons Lab U.S. Army Missile Command (USAMICOM)U.S. Army Armament RDE Ctr Attn: AMSMI-RD-CS-R, DocumentsAttn: SMCAR-CCB-TL Redstone Arsenal, AL 35898-5400Watervliet, NJ 12189-4050 Commander

Night Vision & Electro Optics Lab U.S. Army Tank-Automotive CommandAttn: C2NVEO-RD-NV-GSD, T. Witten Attn: ASQNC-TAC-DIT, Tech InformationFT Belvoir, VA 22060 Warren, MI 48397-5000

Commander USACECOMU.S. Armament RDE Or Attn: AMSEL-RD-EW-R, J. BorowickAttn: SMCAR-TDC (2 copies) FT Monmouth, NJ 07703Attn: SMCAR-FSP-A1, M. Roenbluth Naval Surface Warfare CtrAtn: SMCAR-FSF-BD, L. Yung Attn: Code F41, D. MarkerPicatinny Arsenal, NJ 07806-5000 Atnn: Code F41, L. Mornissett

U.S. Army Field Artillery School Dahlgren, VA 22448-5000Attn: ATSF-CSI Air Force Armament LabFT Sill, OK 73503-5000 Atn: WoIAmnOI

Commandant Eglin AFB, FL 32542-5000U.S. Army Infantry School CommanderAttn: AISH-CD-CSO-OR Naval Weapons CenterFT Benning, GA 31905-5660 Ann: Code 3954,S. Ghaleb

Commander China Lake, CA 93555U.S. Army Misie CommandA.S. ArM issMileD-AS- C m ChriDrexel University Physics Dept

Attn: T. W. MercerRedstone Arsenal, AL 35898-5253 32nd & Chestnut Sts

Director Philadelphia, PA 19001US. Army TRADOC Analysis CommandAttn: ATRC-WSRWhite Sands Missile Range, NM 88002-5502

19

Distribution

Georgia Tech Rockwel Interaonal CorpAtn: M. T. Tuley Attn: K Hull, DC-04GTR/MAL, CRB-564 3370 Miraloma AveAtlanta, GA 30332 Anaheim, CA 92803

Renelaer Polytechnic Institute Science & Tech AssocPhysics Dept Attn: A. GlasserAttn: C. Ventrice 4001 Fairfax Ave, Ste 700110 8th St Arlington, VA 22203Troy, NY 12180 Technology Service Corp

Stens Institute of Technology Attn: B. GrazianoBECS Dept 6515 Main StAtn: S. H. Smith Trumball, CT 06611Hoboken, NJ 07030 US. Army Research Lab

University of Nebraska Ann: AMSRL-OP-SD-T, Tech LibraryElectrical Bug Dept (3 copies)Attn: R. Narayanan Attn. AMSRL-OP-SD-TM, Mail & RecordsMail Stop 0511 MgmtLincoln, NE 68588 Attn: AMSRL-OP-SD-TP, Tech Pub

Balanced Technology Init Attn: AMSRL-SS-I, Chief

An: P. Kcos Attn: AMSRL-SS-F, Chief

1901 N. Beauregard St Attn: AMSRL-SS-SC, Chief

Alexandria, VA 22311 Ann: AMSRLSS-SD, CDiefAttn: AMSRL-SS-SD, D. Bauerle

Coleman Research Corp Attn: AMSRL-SS-SD, D. WignerArm C. Barrett Attn. AMSRLSS-SD, S. Stratton3950 Lakehurst Dr Attn: AMSRL-SS-SD, T. BurchamOrlando, FL 32819 Atn: AMSRL-SS-SD, T. Pizulo (10 copies)

MIT Lincoln Lab Atn: AMSRL-SS-SG, Chief

Attn: S. Ayas, Rm B-.364 Attn: AMSRL-SS-SK, Chief

Atn: L G. Fleischman, Rm B363 A AMSRL-SS-S., ChiefAtnn: AMSRL-SS-SL, Chief

An: P. Reynolds, 2Attn: AMSRL-SS-S, Chief

244 Wood St Attn: AMSRL-SS-S, Chief

Lexington, MA 02173 Attn: AMSRL-SS-S, DirectorAttn-. AMSRLSSA, Chief

Rockwell International Corp Attn: AMSRL-SS-SD, F. BurkeMMW Radar Technology Attn: AMSRL-SS-SH, ChiefTactical Syst Div Attn: AMSRL-SS-SH, J. NemarichAnn: R. H. Wright, MS-DD46 Attn: AMSRL-SS-SJ, Chief3370 Miraloma Ave, PO Box 3170 Ain. AMSRL-SS, J. SattlerAnaheim, CA 92803-3170

20