AD-AL14 117 FEDERAL AVIATION ADMINISTRATION TECHNICAL CENTER ATL-ETC F/6 1/3CONPENDIUM OF LISHTNINO EFFECTS ON FUTURE AIRCRAFT ELECTRONIC- ETCIU)

UNCLASSIFIED DOT/FAA/CT62/30 M

DOTFA/C8230A Compendium of LightningEffects on Future AircraftElectronic Systems

Nickolus 0. Rasch

February 1982

Compendium

This document is available to the U.S. public- through the National Technical Information

Service, Springfield, Virginia 22161.

DTIV7fC) - A 18

US Deparment of Transpotatin ATUchnical CenterAtlantic City Airport, N.J. 08405

-- -- A.

NOTICE

This document is disseminated under the sponsorship ofthe Department of Transportation in the interest ofinformation exchange. The United States Governmentassumes no liability for the contents or use thereof.

The United States Government does not endorse productsor manufacturers. Trade or manufacturer's names appearherein solely because they are considered essential tothe object of this report.

1

Technical Report Documentation Page1. Report No. 2. Government Accessoon No. 3. Recipiont's Catalog No.

DOT/FAA/CT-82/230 -oeen AJesi

4. Title and Suboole 5. Report DateFebruary 1982

A COMPENDIUM OF LIGHTNING EFFECTS ON FUTURE AIRCRAFT 6. Performing Organization CodeSYSTEMS

8. Performing Organization Report No.7. Authorls)

Nickolus 0. Rasch ACT-3409. Performing Organization Name and Address 10. Work Unit No. (TRAIS)Federal Aviation AdministrationTechnical Center 11. Contract or Grant No.Atlantic City Airport, New Jersey 08405 182-340-100

13. Type of Report and Period Covered

12. Sponsoring Agency Name and AddressU.S. Department of Transportation FinalFederal Aviation Administration November 4-8, 1981Technical Center 14. Sponsoring Agency Code

Atlantic City Airport, New Jersey 08405

15. Supplementary Notes

16. Abstract

This publication is a composite of presentations given at the NASA-Langley ResearchCenter/FAA Technical Center "Lightning Effects on Future Aircraft Systems Workshop"held on November 4-6, 1981, at the NASA-Langley Research Center Facility.

The presentations encompassed the full spectrum of lightning research from lightingphenomonology, lightning modeling, electromagnetic issues associated with compositematerials, to the lightning/aircraft electromagnetic interaction analysis. Alsoincluded are a total of five presentations assessing the Digital System upsetphenomen. <

17. Key Words 18. Distribution StatementTriboelectrification Document is available to the U.S. publicCorona through the National Technical InformationStepped Leader Service, Springfield, Virginia 22161Tortuous Channel

19. Security Classif. (of this report) 20. Security Clascif. (of this page) 21. No. of Pages 22. Price

Unclassified Unclassified 251

Form DOT F 1700.7 (8-72) Reproduction of completed page authorized

till III

IAI

A

I' I'

Eu E~

II

2c wo a 0 Q0

* 4

*m

Wa

C2

PREFACE

The "Lightning Effects on Future Aircraft Electronic Systems" workshop,November 4-6, 1981, sponsored by the NASA-Langley Research Center inconjunction with the Federal Aviation Administration (FAA) Technical Center,provided an ideal vehicle for information exchange. This workshop providedregional and headquarters personnel with an insight into the magnitude of theproblem, and the progress being accomplished through existing and futureFAA programs.

protection of electrical and electronic subsystems and equipments against

the atmospheric electricity hazards constituted by lightning and staticelectricity must be taken into special account in the design of advancedtechnology aircraft. Two primary factors have contributed to an increasedpotential hazard to new generation aircraft: (1) The increasingly widespreaduse of digital microelectronic subsystems and/or avionic equipment which areinherently susceptible to upset and damage caused by electrical transientsto implement flight and mission critical functions; and (2) the reducedelectromagnetic shielding provided by many advanced structural materials.Present military and civil design guides and standards are being reviewed toassure adequate protection for new generation aircraft.

The NASA-Langley Research Center Electronic System Branch of the FlightElectronic Division was the host for this workshop; which is a major elementin the FAA Technical Center's Advanced Integrated Flight Systems (AIFS)program. The AIFS program objectives are to acquire and disseminate data,enhance communications, and provide Aviation Standards, lead and/or certifyregions airworthiness/certification personnel with a vehicle for informationtransfer in this highly technological area of lightning research as relatedto aircraft flight safety.

The personnel from the NASA-Langley Research Center Aircraft Electronic SystemBranch and their associates exhibited professionalism in the planning, assemblingthe technical experts and material, and conducting this workshop. It is witha deep and sincere sense of gratitude that we of the FAA would like to extendour appreciation to those persons for a job well done.

-a -tV L Cotinl s

IvalobllitY C'odes

iii

-- : .. . . . .. : _- ., .

TABLE OF CONTENTS

Page

Lightning Effects Research Program - by Mr. Felix Pitts I

Lightning Phenomenology - by Dr. M. Levine 5

Flight Experiment Definition and Summary - by Mr Felix Pitts 19

Data System Description - by Mr. Mitchel E. Thomas 29

Electromagnetic Sensors for Aircraft Lightning Reseach -by Mr. Klaus P. Zaepfel 39

Lightning Modeling - by Dr. M. Levine 51

Interpretation of In-Flight Test Data Approach, Problems and Outlook -

by Dr. R. A. Perala 61

Corona and Streamer Effects - by Dr. R. A. Perala 76

An Analysis Method for the F-106 Direct Strike Data -by Dr. T. F. Trost 81

An Overview of the Electrical/Electromagnetic Impact of AdvancedComposite Materials on Aircraft Design - by Dr. John C. Corbin, Jr. 91

Structural Application of Composite Materials and the DirectEffects of Lightning Strikes - by Mr. William E. Howell 155

Lightning Interaction Analysis - by Dr. Karl S. Kunz 171

Intermittent/Transient Faults in Digital Computers -by Dr. Gerald M. Masson 195

A Microprocessor-Based Upset Test Method - by Ms. Celeste M. Belcasto 213

Diagnostic Emulation Analysis--Need and Techniqueby Mr. Gerard E. Migneault 221

Digital System Hardware Description Technique Used in Emulation -

by Mr. Robert M. Thomas, Jr. 231

The Need for Transient Data in a CARE III Reliability Analysis -

by Mr. Salvatore J. Bavuso 241

v

i ,- "

LIGHTNING EFFECTS RESEARCH PROGRAM

by

Mr. Felix Pitts

Langley Research CenterNational Aeronautics and Space Administration

The current research program being pursued at the LangleyResearch Center on lightning and its effects on digitalelectronic systems will be presented.

4 -

A-AA

-I

LIGHTNING EFFECTS

CURRENT TRANSPORT AIRCRAFT STRUCK ABOUT ONCE PER YEAR

- ALUMINUM ,SKIN

- HYDRAULIC/MECHANICAL PRIMARY CONTROLS

- MOSTLY NUISANCE PROBLEMS CAUSED BY LIGHTNING

m FUTURE AIRCRAFT WILL EMPLOY- COMPOSITE STRUCTURE

- DIGITAL AVIONICS/ELECTRONIC CONTROLS

SUSCEPTIBLE TO DISTURBANCE

POTENTIAL FOR UPSET

. EED FOR FUTURE AIRCRAFT DESIGNS

- BETTER UNDERSTANDING OF IN-FLIGHT LIGHTNING ENVIRONMENT

- TECHNIQUES FOR ASSESSINIG DIGITAL SYSTEM PERFORMANCE INLIGHTNING ENVIRONMENT

LIGHTNING EFFECTS RESEARCH PROGRAM

D SA/C RESPONSE:SIUMENTATION SYSTEM/CALE DIAGNOSTIC EMULATIONli .c ....iI/VOLTAGES AD -STATE TRANSITION

ILIHTNING SOURCE: CURRENTS

I DIGITAL SYSTEM "

~USET ITEST TEST DNFINIT

3

LIGHTNING PHENOMENOLOGY

by

Dr. M. LeVine

Goddard Space Flight CenterNational Aeronautics and Space Administration

State-of-the-art of lightning knowledge: Theories ofcharge generation, the lightning discharge, steppedleaders, return strokes, fundamental electromagneticsof lightning, statistical distribution of lightningcharacteristics.

5 fwmium p~aa u nam

10

+ -33

CI..OUD - CLOUD -8c

CLOUD - GROUND

0 +30

70psec 60ILS

L~\ Mscim

h -20 msec ~ 440 mriec-P, I.4 30 mseci>I j60pnc

DART DARTLEADER LEADER I

-~:STEPPEDv

LEADER

RETURN RETURN RETURNTROKE STROKE STROKE

LU t

to r- c LU0 0 0I~

LU 24

LC0 0o 0- CI-,wl 023 c-~o .- 0

> 4 a r

LU otL

LU uj x x

2U 00 R cc o=L to N~' La Lo C04 U-(VU-( ACD a

I..

c'- oo LU o U0f WL44 4jW)r e(

0O 0 0 us CC

~80 ~ 0 ~OL4 x x C4O > Q t- i4

Mi oJ C!JZlb WLU

C-3 in N -a -

cm to c

HISTORICAL PERSPEC VE

1752 B. FRANKLIN

IM6 FIRST PHOTOGRAPH

1916 C. T. R.WILSON(FIELD MIWL

1929 BOYS AMERARADIATION RELD(OS 1 = INS

STUDIES ON OUTH AFRICA

SCHONLAND

1980 STERD PHOOS) NEW MEYJCO TECH GET CA o

WIDEBAND DIGTAL EMI TI 1170I

THUNDESTORM RESEARCH PROJCT

Ism. Is

maI

TU

UC

'.4

E9

>49

10

CL

C I-~ch

LLU

I -

LLU

LU

cowL

DISTRIBUTIN OF CHAR85 IN A THUNDERCLOUD

+ + -45

113

FIELD HANGES3STE

00

AN4TENNA INTEGRATOR LOAD

E I NTEGRATOR

12

SLOW ELECTRIC FIELD CHANGES

INThAGWUD

CLOUD -GROUND

FAST ELECTRIC FIELD CHANGES

LOW RESOLUTION HIGH RESOLUTION

FIRST RETURN STROKE

SLOW E

CLOUD PROCESS

FAST E

I A -

0 .1 .2 .3 .4 0 20 406080 100

TIME (SECONDS) TIME (pb

13

FIRST RETURN STROKE

SUBSEQUENT RETURN STROKE

INTRA CLOUD PROCESS

I I I I I I I I

-400 -200 0 200 400TIME (/.Ls)

14

* S - t

RADIO RECEIVER

a.OuO-cOUO RAS

CLOu-mUN RASH

0. 01 . 0.3 0.4 & .0 07 0.8 0.89 110 1.1 I,,TIME WSCONOS)

3 MHz 2~~

30 MHz

139 MHz

295 MHz

SLOW E

0.0 0.5 1.0IME (SECONDS)

15

-~ . a7

3 MHz

30 MHz

139 MHz 1. .

295 M~I

!

SLOW E

I I I I I I I J

0.0 0.5 1.0TIME (SECONDS)

OPTICAL

SLOW E

FAST E+

RF 1W MHz)

. 0 w 100 160 20 2W

TIME (MILUSECONOl

416I .-

MAPPING USING RF RADIATION

A - -- - - -. --3 *

1 t2

-2 00 202

30SUBSQUET RETURN

STROKErz15-

I w

0 20 40

157

17I~

30

25

20 -I(T 0[aO - *aJJT) + OY-0-T

~1a - 20104 SeC- 1

uj 15 -5015 ec-,cc ~y - X10 3 SOC- 1cc

5

0 I0 20 40 60 80 100 120 140 160 180 200o

TIME (us)

18

FLIGHT EXPERIMENT DEFINITION ANID SUMMARY

by

Mr. Felix Pitts

Langley Research CenterNational Aeronautics and Space Administration

Description of NASA F-106B in-flight direct strike measure-ment program. Electromagnetic measurements, instrumentationconcept, and results summary.

19

SI*PIJF)ED LiGHTING MODEL

ELECTROSTATIC r,\~" CAG

H a (tDc a it-0/ +i a it-D/ci

I.H + + ID.

20

MEASUREMENTS SUMMARY

AMPLITUDEMEASUREMENT SYMBOL RANGE SENSOR TYPE

RATE OF CHANGE OF FLUSH PLATE

ELECTRIC FLUX DENSITY 50 A/M2 DIPOLE

RATE OF CHANGE OF MULTIGAP

MAGNETIC FLUX DENSITY B 2 X 104 TESLA/SEC LOOP

RATE OF CHANGE OF INDUCTIVE

CURRENT 10 ' A/SEC CURRENT PROBE

MEASUREMENTS SUMMARY

MEASUREMENT SYMBOL DIMENSION SENSOR TYPE

* RATE OF CHANGE OF FLUSH PLATEELECTRIC FLUX DENSITY t A/M2 DIPOLE

* RATE OF CHANGE OF MULTIGAPMAGNETIC FLUX DENSITY TESA/SEC LOOP

* RATE OF CHANGE OF INDUCTIVECURRENT I A/SEC CURRENT-PROBE

* ELECTRIC FIELD E V/M FIELD MILL

* CURRENT I A CURRENT TRANSFORMER

21

MEASUREMENT LOCATIONS - 1980

D----(ELECTRIC FLUX DENSITY)

=A (MAGNETIC FLUX DENSITY)

A= - (TOTAL CURRENT)

40 x 103 1980 LIGHTNING OCCURRENCES

69 PENETRATIONS 10 STRIKES

35-

30-

25 -

ALTITUDE

20FEET

1 .5

4 5-

0 l1b 2 0 2 4 6 .5 5.0

PENETRATIONS STRIKES STRIKES/PEN

22

-~Mil

INSTRUMENTATION SYSTEM

EXPNDI!DBI)MTIO

RCA DVISR TRNSIET REORZE

23mv-

24

I

- I

fiSi

C

L C --~ a C.

C - C

I a,'Ii'

* a a qaq.q.eea.q~..q. I b.

333 ~ i * !!!~ 3 ~ U,

* ,.,v mu is-a L.minII.diI..iiiiiiiieg~

a a ~ * U q aI-

~* m4

I--J

cd~

ESC

C

I- I -

C.D- B - a --J - ~ I,

- L U.LE ~.__ - - a C- E C A

I :~ aI- - Sci-,I- IC -

L.i..J - ~ pill, liii

* Si- U .3 iiitIiIII!tIIIIIIIIIif ! ~ meT

.- ib muiaoa sj ljgg

a-I-

I

-C

CU,

Si

.1 >1.4 -~= a,

Ii* - m

IC

a

* a a - ~

.*,u muicco 3~1 1AfII~

25

1981 LIGHTNING OCCURRENCES

40 ),13.

Ill PENETRATIONS 10 STRIKES

30

ALTITUDE 2

FEET

103

00 10 20 0 2 4 0 1.0

PENETRATIONS STRIKES STRIKES/PEN

TIME: 19 18-39 DATE AUGUST 9. 1901LOGGER NUMBER: F 106 FLIGHT NUMBER: 43

+25000 - -

20000 -

1500 - -- -

10000 -

4020000 60 20

TIME (MICROSECO#405)

X INDICATES TRIGGER POINT

26

DATA SUMMARY

1980

- 19 FLIGHTS 69 STORM PENETRATIONS

- 10 STRIKES 17 TRANSIENTS: (7 b, 5 t, I)

- RESULTS REPORTED:NASA TM 81946 "1980 DIRECT STRIKE LIGHTNING DATA'AIAA 81-0083 "ElM MEASUREMENT Of DIRECT LIGHTNING STRIKES TO A/C"

24 FLIGHTS 111 STORM PENETRATIONS

10 STRIKES 27 TRANSIENTS

1 BOOM CURRENT (BOEING)

16 t, 10 6 (DISTANT)

MAXIMUM MEASURED VALUES

b 30.5 AIM2 > 340 KV/O,1 ,s

8 1160 T/s - > 2 KA/O,I ,s

I 15 KA

OTHER F-106 LIGHTNING EXPERIMENTS

e ATMOSPHERIC CHEMISTRY EXPERIMENTS - LARC

1980 - SHOWED N20 ENHANCEMENT NEAR LIGHTNING

1981 - SHOWED CO ENHANCEMENT NEAR LIGHTNING

o X-RAY EXPERIMENT - U. WASHINGTON

1980 - SHOWED INCREASED COUNT NEAR LIGHTNING

1981 -BEING EVALUATED

a OPTICAL SIGNATURE EXPERIMENT - NSSL

- PROTOTYPE FOR ORBITAL LIGHTNING MAPPER

- RESULTS BEING ANALYZED AT NSSL

@ STORM SCOPE-TURBULENCE/LIGHTNING DO NOT ALWAYS CORRELATE

- LIGHTNING LOCATIONS DISPLAYED TEND TO BE FURTHER FROMAIRCRAFT POSITION THAN RADAR CONTOURS (BUT AT SAME BEARING)

o BOEING DATA LOGGER

- FIRST BOOM CURRENT RECORDED AUGUST 1981

27

DATA SYSTEM DESCRIPTION

by

Mr. Mitchel E. Thomas

Langley Research CenterNational Aeronautics and Space Administration

The research data-gathering system on the F-106B aircraftdeveloped for in-flight measurement of direct and nearbylightning strike characteristics is described. Details ofthe design and performance are presented for system com-ponents including the digital transient recorders, widebandanalog recorder, fiber optic control and diagnostic links,power system isolation, and system shielding.

.

29

It M05fl PA(Z DLaaS.J ium

L4 "

LIGHTNING INSTRUMENTATION SYSTEM

o WIDE BANDWIDTH FOR SUBMICROSECOND SIGNAL RESOLUTION

- 10 NS SAMPLE INTERVAL

* CONTINUOUS RECORDS FOR DEFINITION OF FULL LIGHTNING SCENARIO

- 15 MHz ANALOG BANDWIDTH

s PROTECT AGAINST SPURIOUS RESPONSES (EMI)

- SHIELDED ENCLOSURE

- MOTOR-GENERATOR POWER ISOLATION

- FIBER OPTIC CONTROL

s AUTOMATIC OPERATION

UGHINING INSTRUMENTATION SYSTEMSYSTEM CONTROL-'_ f'-IRIG B TIME

IFIBER OPTIC N I

SCURRENTI EE ELSMHz

, I 1 MPLFE ANALG

BOOM 12 2C)RCRELOOPS TRANSIENT G;

• FUSELAGE RECORDES& WINGS 4)(2 CH)

FWD.. TA IL, ITGAO'16 CH) MGEI

&WINGS 11TP

"RcR''- :' Z ZGE"ER'OR H RECOR'DRE -_PWFIELEDENLOUR

(14!C

I '~ ~ INSTRILmENTATioN SYSTEM

- *EE~EXPNED EUUUUEUEEUEUEE~~~~NAumEC RECORDERuupuuuuummuuuum

RCA DVISR TRLIE EU U4- UPOE SUPPLI 200

ELECTRC MOTO

TRANSIENT RECORDER

6-BIT 131072 6-BIT-0AID WRD/AANALOGCONVER- WODDIA i

MCONVER- CONVER- OUIPUSPTE TER

DIGITALOUTPUTS

S IGNAL

INPUT

INT. • TRIGGER, TIME BASEEXT. TRIGGER __1_AND

INPUT EXT. CONTROL CIRCUITRY

EXPANDED TRANSIENT WAVEFORM RECORDER FEATURES

I 6-BIT AMPLITUDE RESOLUTION (1.56%)

1 FREQUENCY RESPONSE - DC TO 50 MHz

* 131072 (217) DATA WORD CAPACITY

I SAMPLE RATES UP TO 100 MHz

I DATA WINDOW LENGTH OF 1310 MICROSECONDS AT 100 MHz

I INTERNAL OR EXTERNAL TRIGGERING

I DATA WINDOW SELECTION WHICH IS BEFORE, DURING, OR

AFTER TRIGGER EVENT

32

33

WIDEBAND ANALOG RECORDER

o VIDEO RECORD TECHNIQUES

s FREQUENCY MODULATED CARRIER

o 10 Hz TO 15 MHz BANDWIDTH

a 12 MINUTES RECORD TIME

* 2 CHANNELS & 2 AUXILIARY

LOGARITHMIC AMPLIFIERS 22222 .L s 1v

161o 11l. 1. 8K l022 23

l6pf 47lO47f

• IS

S11A2 6. 68~ 56.f-- 6111 022

IO 8 222 T-- is8 0

lofIf 2N97l lOOf .

3.5 ' 22 ;43

,! 22

34

COMPRESSION AMPLIFIER

s 15 MHz BANDWIDTH

s DIFFERENTIAL INPUT, 50 OHM IMPEDANCE

o E0 - 1 + .3 LOG EIN

o 60 dB DYNAMIC RANGE

1.0 -

INPUT o.i -VOLTS

0.01 -

0.001 -1

I I 10 .2 .4 .6 .8 1.0

OUTPUT, VOLTS

SIMPLIFIED SHIELDING TOPOLOGY

AIRCRAFTMETAL SENSRSURFACES

7

rDATA SYSTEM DATA RECORDERMETAL ENCLOSURE CHASSIS

i I APERTURE PENETRATION

35

LIGHTNING INSTRUMENTATION CONTROL

FBRFIBER ONE POE

POWIR ON/OFF SW. = OPTIC - OPTIC DEODR CONTROLXMRRCVR [

FBRFIBER] TONE ITAPE I

TAPE RUNISTOP SW.- - - -OPTIC DECODER CONTROLXMTR RCVR

RECORD TONE FFIBER FIBER TN

INDICATOR DECODER OPTIC - - -- OPTIC REC/w-Ec TRANSrIENTRCRXMTR RECORDER

RECOAD TONE F~IBER FIBER TONEINDICATOR& DECODER OC --- PI REC/v~ TRANSIENT

RCRXMTR RECORDER

RECORD TONE FIBER F IBE R ,-..TONE MAGNETIC

INDICATOR DECODER OPTIC - - - OPTIC TAPERCRXMTR X1 REC/w RECORDER

COCKPIT FIBER SHIELDEDCONTROL OPTIC ENCLOSUREPANEL CABLES

x109

6 I-DOT SENSOR RESPONSE

04z0L)

WN,

-6 1LLIjIJ 11 fi uiujjuju u ill]550 S75 600 625 650 6.75 700

MICROSECONDS

36

LU

V) 4;

04

z4

373

ELECTROMAGNETIC SENSORS FOR AIRCRAFT LIGHTNING RESEARCH

by

Mr. Klaus P. Zaepfel

Langley Research CenterNational Aeronautics and Space Administration

Electromagnetic sensors designed for measuring EM fieldsand currents during lightning strikes to the F-106B researchaircraft are described. Sensor theoretical basis, perform-ance, and parallel-plate transmission line used to check andcalibrate these sensors are described.

39

PMEUING ?AMu BLAUE.iT FIUW

MEASUREMENT REQUIREMENTS

MEASUREMENT FIELD CHANGE OUTPUT

1 50A/m2 (A E ~ 6 x 10 _ per 0. ips) > 100V

2' 2xle Teslals > 100v

IlOkA per 0. lp s > bOyv

SENSOR CHARACTERISTICS

*DESIGN PRINCIPLES, DEVELOPED BY AIR FORCE WEAPONS LAB FOR NEMP

*WIDE BANDWIDTH: >:80MHz (tR :S 44 fls)

*SIMPLE GEOMETRY: CALIBRATION BY RULER

*APPROXIMATELY 100 V OUTPUT FOR FULL-SCALE DIRECT STRIKE

40

DEFINITIONS: ELECTRIC FIELDS

E.-FIELD , vl, IC CORONA(Ekedric Field Irbicisity) +CULRRENT

IC It i- D-Area- 0DIPLACEMENr

I t TOTAL CURRENT

NoCroncL, rt- 6 AreA

D, C/-- In Coronk, Its (Jo' 5)AreaEkedrkc Flux DeA sty) trL

(Surfctcw-, Charje Dwisity)

Forfla pte, D--ps

D- c E. wh4ere f 't

0tevr.tivitr of dkctic.

SENSOR FUNDAMENTALS

CLOUD or LIGHTNING

CHANGING E-FIELD(it)

AI RCRAFTA FUSELAGE

NO CORONA CORONA

it.I4 D5AREA it-{ct ) AREA

[4 GROUND 1VOUT, R It

41

£ .. -- -

D-DOT SE -NSOR IN F-106 FUSELAGE PANEL

42 iD A

'JT SENS011

/lo/ 1 =2.2 RC, R 5W0/ rrl r t = tr+ PROPAGATION TIME

rj A

~*,out i T

\\ \ i'4" -~A =irr2, r2 rr

-'4 1 12

SECTION A-AT

DEFINITION: MAGNETIC FIE.LDS

/

MAXWELL-AMPERE LAW: dlfJA

+2-1Tr IAAr

43

SENSOR FUNDAMENTALS

El. -A-I

MAXWELL-FARADAY LAW: ACd -. dA

V 4 dA B- DOT S.,.,~

cit fJA2rf

v._,.-Mi, M-?,LIA-dA I -DOT Seuisor

B-DOT SENSOR

B

Vot 2. -R 10 1000A00 -1r2 100

B2v OUT

44

[B PWU I~ Nl.)R ON F- V(b F U 0 I A(,[ PANEL

45

I-DOT SENSOR

10Q(L s+ 1.t r 22 R

100 vout m

10 -21T b

OUT ~~CO t4D UC TORTO A IRFRAME

SENSO GROUN PLN

TeeTi 44

GENERDUCTOR

FLTPAETA SMION LINEGCLRATORDSPA

OSCILLOSCOPEPE OCLLOSOP

E4

BO' SESO GR-N PLANE--- ~~.

Tee_ _ _ _ _ _

PU--L- -SFA

FLAT-PLATE TRANSMISSION LINE CALIBRATOR

7CALIBRATOR

- "-' ; . 11 ROUND PLANE

47

Lu

LnU

LALU

LUJ LUJ

Il) >

(U)

Wo0Ur.

C%A

U)C

WlI--

0

E )

48

SUMMARY OF SENSOR PARAMETERS

SENSOR LOCATI ON DESIGN SENSITIVITY RISETIME

i - FUSELAGE/TAIL CIRCULAR 4. lx10_2 2 2.4n

STWINGS RECTANGULAR 2.74102 m 2 3.0 ns

B-DOT FUSELAGE 2 GAPS, 4 LOAD PO INTS 5703m 2 .85 ns

B-DOT WINGS I GAP, I LOAD POINT 5.54103 m 2 2.0Ons

I-DOT RADOME 1 GAP. 4 LOAD POINTS 2.l4xl9 H 0.8 ns

FIELD MILLS

Ev

49

mks

50

LIGHTNING MODELING

by

Dr. M. LeVine

Goddard Space Flight CenterNational Aeronautics and Space Administration

The ground-based lightning measurements made at WallopsFlight Center concurrently with the in-flight direct strikemeasurements are described. Lightning mathematical modeldevelopment to arrive at credible lightning models for use

in induced effects electromagnetic coupling studies isdiscussed.

51

i, : .. . .. .. .... . . .. .. . . .... a w

AL

...... OBSERVER

OBSERVER

IMAGE

CHANNEL GEOMETRY

X-Z PLANE Y-Z PLANE4

3-

2-

0

2 1 1 2 2 1 1 2

* KILOMETERS

52

CHANNEL GEOMETRY

XZ PLANE V-i PLANE

6-

5

* 4-

3-

2 -

01

-3 2 -1 1 2 3 -3 .2 .1 1 2,3

KILOMETERS

30

25

20 -I(T) 10 I(e -T - e-l)+ j(e-yT .. 6 -dT) 1

I- = 2004 sec-1uz15 f3=51 5 sec-I

cr- y - 1010 3 sec-1

o 10 d = 20104 SeC-1

5

0 - 10 20 40 60 80 100 120 140 160 180 200,

TIME (ps)

53

CHANNEL GEOMETRY0401 ,6 .6

6.000. Y.Z PLANE 6.000 X-Z PLANE

J4' 4,000

S2.000 2.000

1,000 1.000-

.400 -3375 220t0 .1125 0 1120 2200 -22W0 .1126 0 1121 2W0 3375 4W0METERS METERS

I .146 SPECTRUM OFRADIATED WAVEFORM

V-RADIATED WAVEFORM 20 LOG IE(vl'

,0 2 4 oo S I00 10. Ile lo e lop i too

TIME IMlefflconmd PlIOUINCY (HE)

>

w

- SIMULATED L SIMULATED

04 s o020 40 60 80 100

TIME (Microweconds) TIME (Mkerwconds)

LL MEASURED 14MEASURED

S 0 20 40 60 60 10 40 m 1t0oN

TIME (Microseconde) TIME (Mkomnnds)

54

00

00

00

-ISO0 o HORNER & BRADLEY, 13'640x KIMPARA, 1965

-200 I

1 10 102 og104 lo0 log

FREQUENCY 1kHz)

1.0

0.6-

0.2

-0.20 20 go0 3 100

TIME (IA a)

-70-

-110

-130110 102 I03 104

FREQUENCY (kHz)

55

-40

wm

-5s0

LVL

-2001.01 10.10 103 10.0

FREQUIENCY (Hz)

o 56

A-L

RADIATED WAVE FORM X-Z PLANE Y-Z PLANE

4-

2-

0-4

.4 3

S 0 a0 40 D S0 0100 -2 -1 1 2 -21 1 2

0 3

-2

~f 0 20 40 60 so0100 -2.-1 1 2 -2 -1 1 2

4-6

2

-2 .

.4

0 20 40 60 6 100 -2 -1 1 2 -2 1 1 2TIME fps)I KILOMETERS

-10-TORTUU CHANNEL

:) -150-STRAIGHT CHANNEL

-200

1 010.031) 0

FRQECwkz

C57

cr.150

wa-

Cd) THEORY

SERHAN, ET. AL. (1980)

---- LeVINE & MENEGHINI

-200

1 10 102 103 104 105FREQUENCY (MHz)

. . .........

020 40 so UD

TIME tWSI

1.1061

imsim

59

cLOSER

- - ~p - - - - - - - -

I -p -II -II /I /I /

//

//

/I /I /I II /I II I

II /I II I

II

I* I I

I~

I4 Il

'4~

0 4 8 12 *16 3

TWE (M 8)

60

r-r---r fl

INTERPRETATION OF IN-FLIGHT TEST DATA APPROACH,PROBLEMS, AND OUTLOOK

by

Dr. R. A. Perala

Electromagnetic Applications, Incorporated

Discussion of the direct strike data interpretation pro-blem and issues to be resolved. Review of lightning/aircraftinteraction process and aircraft electrical resonance con-siderations. Approach for generalization of F-106 data toother aircraft classes.

61

OUTLINE

I 11 _16H1NINI~bA RCAI I INIL.RALI ION PRO.1Y,

I THE NEED FOR IN-FLIGHT TEST DATA

I EXTENSION OF DATA TO OTHER AIRCRAFT

I LIGHTNING MODELING

I AIRCRAFT MODELING

I EXAMPLES

STEPPED LEADER APPROACHING AIRCRAFT RETURN STROKE THROUGH THE AIRCRAFT

STEPPED LEADER ATTACHMENT AND CONTINUED NO RETURN STROKE THROUGH THE AIRCRAFTPROPAGATION FROM AN AIRCRAFT

62

A * -

LIGHTNING/AIRCRAFT INTERACTION DESCRIPTION

* AIRCRAFT IN LARGE STATIC FLECIRIC Fifil), 10-100 (V/M IYPICAl

A AI RCRAVI rIOuIABLY At. RI AI)Y C IiAli( II IU I I I I II I ( 11, 111 AIItN

* APPROACHING STEPPED LEADER GIVES LARGE E, 11/'I-

* WHEN E LARGE ENOUGH, AIRCRAFT STREAMERS

* STREAMER ATTACHES TO APPROACHING LEADER

* CHARGE FROM LEADER DEPOSITED ON AIRCRAFT AND ELEVATES ITS POTENTIAL

* WHEN ITS POTENTIAL IS HIGH ENOUGH, LEADER CONTINUES FROM AIRCRAFT

TO DESTINATION

* STEPPED LEADER CURRENT FLOWS THROUGH AIRCRAFT UNTIL RETURN STROKE

* RETURN STROKE CURRENT LOWERS AIRCRAFT POTENTIAL

* AIRCRAFT IN CORONA DURING MUCH OF THIS TIME

APPROACHING LEADER E FIELDS

6000

400

. 00

0 100, 400 600 60go0o0

DISTANCE M HEIGHT ABOVE GROUND M

ELECTRIC GRADIENT BELOW LEADER CHANNEL: (a) AS FUNCTIONOF HORIZONTAL DISTANCE AND (b) AS FUNCTION OF HEIGHT OF

LEADER TIP ABOVE GROUND

ASSUME DE . Of lh ?h 6 nSe

n T IT and iT .5x10 se

WE GET Ki - 1.2 x 10atOO000at

1.5 x 1010 at 10-20 111

63

AIRCRAFT STREAMERING

CURRENTS PROBABLY LESS THAN IOOA, BASED ON GROUND STREAMER DATA

* WAVESHAPES: 10 vsec RISETIMES

* NEGATIVE CHARGE ENTERING AIRCRAFT AT STREAMER POINT

* STREAMERING CAN OCCUR FROM CRITICAL POINTS

I STREAMERING INFLUENCED BY STATIC FIELD

I ELECTRIC FIELD ON AIRCRAFT REVERSES AT ATTACHMENT

STEPPED LEADER CURRENTFLOW THROUGH AIRCRAFT

AIRCRAFT NEEDS TO ACCUMULATE 100 PC OF CHARGE FOR STEPPED LEADER TO

CONTINUE ON FROM AIRCRAFT

* IF LEADER CURRENT iS IO00A RISING IN 100 nso CHARGING TIME IS ABOUT 150 ns

* DURING THIS TIME AIRCRAFT NORMAL ELECTRIC FIELDS APPROACH 3 MV/m

* THIS GIVES E/at 2 x 1013 V/m/sec

" LEADER CURRENT CONTINUES THROUGH AIRCRAFT AND ELECTRIC FIELD STAYS

CONSTANT AND CURRENT ASSUMES THE SLOWLY INCREASING ARC CURRENT UPON

WHICH IS SUPERIMPOSED LEADER PULSE CURRENTS

* AIRCRAFT IS IN CORONA DURING THIS PHASE

64

100-500A

usJ~

STEPPED LEADER CURRENTS

J- AND K- CHANGES

J-CHANGES; SLOW, 100A CURRENTS

K -CHANGES: SEVERAL THOUSAND A~MPS

POSSIBLY 50 Ns RISE TIMES

65

ISSUES OF IN1[RL'I, AND

THE NEED FOR IN-FLIGHT TEST DATA

I ATTACHMENT PROCESS

* WHAT ARE I AND

* HOW IMFUrTANT ARE NUNILINEAIF (tUFFIINA, ';ITRIAFt I 111 1 !v!,

* WFAT IS TIlE EQUIVALENT CIRLIlI F 1i1 IFAlIF (hIFANNLL IMIlIANLlF

NORTON CURRENT SOURCE)

* RETURN STROKE

* WHAT ARE E AND ,_ ?

* WHAT ARE H AND jF?

* HOW IMPORTANT ARE NONLINEAR EFFECTS?

I WHAT IS THE RETURN STROKE EQUIVALENT CIRCUIT?

I NEARBY LIGHTNINGi How IMPORTANT IS IT?

* How APPLICABLE ARE STATE OF THE ART LIGHTNING MODELS AT AIRCRAFT ALTITUTE. "

0 MODELS SO FAR ARE BASED ON TERRESTRIAL OBSERVATIONS

* HODELS HAVE NOT BEEN TESTED AT AIRCRAFT ALTITUDES

I OBTAIN DATA ON INTRACLOUD LIGHTNING

I HOW DOES THE AIRCRAFT INTERACT WITH LIGHTNING? IS TilE LIGHTNING EN ;"

ITSELF MODEIFIED BY THE PRESENCE OF AN AIRCRAFT? IS IT DIFFERENT ,

DIFFERENT AIRCRAFT,

THE OBJECTIVE OF DATA INTERPRETATION

I WHAT ARE THE LIGHTNINGS THAT CAUSE THE F106B RESPONSES?(AN INVERSE PROBLEM)

I WHAT WOULD BE THE RESPONSE OF A DIFFERENT AIRCRAFT?

I CAN A METHODOLOGY BE DEVELOPED FOR EXTENDING IN-FLIGHT DATA

FROM ANY AIRCRAFT TO ANY OTHER AIRCRAFT?

66

' ____ .... __- - --- L 4-- . .,

APPROACH

Lightning F106 F106 1106 Grnd.TesEnvironment 3DFD Computer In-Flight Data (ScalelMModels Model Data deling fl

Checkout

Sai ignfcteoyNnierte

Mdedtffccation of ECraoltinpehoolg

Measur7

Predicted~

IMPORTANCE OF NEAR MISS DATA

I INTERACTION CALCULATIONS CAN BE DONE LINEARLY, WITHOUTCOMPLICATIONS INTRODUCED BY NONLINEARITIES

I DATA CAN BE USED To INFER NEARBY LIGHTNING ELECTROMAGNETICWAVEFORMS AT AIRCRAFT ALTITUDES To CORRELATE WITH PREDICTIONS

/R

/ GROUND (or-0)

a 4R

A DRAWING DEFINING ALL GEOMETRICAL PARAMETERS NEEDED IN TNECALCULATION OF ELECTRIC AND MIAGNETIC FIELDS. (FROM4 UI4AN)

68

Slope v

Z4 V'4

-Unit arm

Current

Z32 Vt3

breakdownPUiWIcurrent

Z22 V12Corona

Height, z curn

12 3 14 oil t2 13 14 rlime

CURRENT DISTRIBUTION FOR THE MODEL OF LIN ET AL. (1980) IN WHICH THEBREAKDOWN PULSE CURRENT IS CONSTANT WITH HEIGHT. THE CONSTANT VELOCITY

OF THE BREAKDOWN PULSF CURRENT IS V. CURRENT PROFILES ARE SHOWN AT FOUR

DIFFERENT TIMES t1THR UJGH t, WHEN THE RETURN STROKE WAVEFRONT AND THEBREAKDOWN PULSE CURRENT ARE AT FOUR DIFFERENT HEIGHTS ITHROUGH z24RESPECTIVELY, (FROM UMAN)

20

-total current---*corona Current

15 uniform current.... pulse current

..... ~ ~ -

0................., -5L 6---. - -- -

10 20 30 40 0 60 70 so 90 100A4 TIME (Pssc)

RETURN STROKE CURRENT COMPONENTS AT GROUND FOR ATYPICAL SUBSEQUENT STROKE CALCULATED FROM MEASURED

ELECTRIC AND MAGNETIC FIELDS (FROM UMAN)

69

.........

......................

ai

A 1zoo

* 3 3 3

- u~m *toits-

70 M

mAlka

0

LA-

a LI

N UUz

2 ac

LL

LU

L-

LU

La

z

TLU

Lt 0

inj

'AMA.

r-4

-AJ

Uj-

72

at (a. c

L)> 1(t.Jiz * t1 CxC)(j).Z(k.yn)l M _______

(''A (JIA is (k) - (W)S t*U). I y ,z

m () * 1-1/)AR * ~ ) * (-1)a y * (k) - (k-I)A.z *f (n- )A t

.A1'an

V z%11) - )u - 71) ) clI -1'

at1 1 -~

S.F.t (11,R) * Is"(Jk- 0 '~l Mbr )lZCiJk) 'ICZ' 2*t /.I - *I)

wck an]

3-912 W(I1*.J.) l-i 1'1tJk 5~(~~) -H"( Jk clk~ 1

'Ii).) 7,u) ( I(1 .1uaI

l cl ean-

all~~~~ ~ ~ y:1L4g~Y4~~tlc~. Jel -,cad(nI

a a .cancan

A. 1.4.k ) *y ."I kI y.J J,,-.~k *c :l (,n-'

M ~ *411 2cl~ kcl.1 Ia a

3-D FINITE DIFFERENCE EQUATIONS IN RECTANGULARCOORDINATES SET UP WITH EXTERNALLY SUPPLIED H FIELDS

73

.36 '0

~ .2 200 2C . -200

15

~-400 10

S-6000

-. 1 -800 150 1 2 3 4 -10000 1 2 3 4 01 M 2 3 '

TIME (is) TIME (wTIE us

DATA FROM FLIGHT 80-038, RECORD 4

20

-MEASURED

15

~10

I ol ;j- - - - - - -

IIt

.1 'I1CJMPUTED

-51I

0 100 200 300 400 500

TIME (NS)

COMPARISON OF MEASURD RESPONSE (FLIGHT 80-18) AND COMPUTED WITH T3DFD ANDA 30Ns RISE TIME, 590 AMPERE STEP CURRENT SOURCE, NOSE TO TAIL.

74

CORONA AND STREAMER EFFECTS

by

Dr. R. A. Perala

Electromagnetic Applications, Incorporated

Description of possible corona effects on direct strike data,review of elements and state-of-the-art of corona modeling,and application of corona modeling to lightning/aircraft inter-action. Corona to arc transition, importance, and modeling of

streamers.

75

75

z0

0:0

LL LUI

0 i0

00

0-0

zZo F-

I- 4I

C:):

L-

I-- 0(

LL.LLU)

F--

0 I-

C.

C:.)

76

CORONA: BASIC MECHANISM

I REQUIREMENT: FREE ELECTRONS ACCELERATED BY ELECTRIC FIELD

Avalanche Electrons

dn - ndxe Ii n - Nunmber of Electrons

Trigger Electron AirMolecule x Distance

\ 8 a -Townsend Ionization Coefficient

Ilectric He~ld Positive

Ion

I' SOURCE OF FREE (TRIGGER) ELECTRONS: BACKGROUND IONIZATION CAUSED BY COSM1IC RAYS;r ..2 X 107 ELECTRON-ION4 PARIS/(m 3 _sec). THIS GIVES AMIBIENT AIR CONDUCTIVITY

10-3 ho/in.

77

L ist 'm.

ELECTRON AND ION LOSS MECHANISMS

* ELECTRON ATTACHMENT TO NEUTRAL AIR MOLECULES TO FORM NEGATIVE IONS

ATTACHMENT RATEE

RECOMBINATION OF POSITIVE IONS AND ELECTRONS TO FORM NEUTRAL PARTICLES

RECOMBINATION RATE 8

* RECOMBINAZION OF POSITIVE AND NEGATIVE IONS

RECOMBINATION RATE Y

AIR CHEMISTRY EQUATIONS FOR CORONA

dnet)t + [o 4 (t) + ae G I n e(t) Q~)

dt- e e

dn,(t) + [Ble (t) + yn_(t)] Q(t) + G ne (t).

dt

*AIR CONDUCTIVITY a - q (p e n e + vi(n. + n+)

I CONDUCTIVITY - FEEDS BACK INTO MAXWELL'S EQUATIONS

A x +

I COEFFICIENTS G, e, 'Ve IN TURN DEPEND UPON E

I COEFFICIENTS ALSO DEPEND UPON HUMIDITY AND AIR PRESSURE

78

* IF WE ASSUME SPACE CHARGE NEUTRALITY

* HOWEVER, SPACE CHARGE EFFECTS ARE IMPORTANT FOR CORONA, SO WE NEED TO INCLUDL:

V 5 + =0

p (n+4 n_ ne q

J = (n+ n n e e) q

ALSO NEED TO INCLUDE PARTICLE DYNAMICS

F s (f + V X x B), s REFERS TO SPECIES

RESULTS FOR A ROD-PLANE GAP.NUMERICAL RESULTS ASSUME SPACE CHARGE NEUTRALITY,

64 KV 3

'40ONS MEASUREDRISETIME L

2-

.15cm

PLANE 0 100 TI 00 300

79

EFFECTS OF NONLINEARITIES

* AIR CONDUCTIVITY MAY SHIELD APERTURES...

IE

AIR BREAKDOWN MAY CAUSE INCREASES IN at o FOR EXAMPLE, AS

INDICATED ON A PREVIOUS SLIDE

* AIRCRAFT COMPLEX RESONANT FREQUENCIES MAY CHANGE BECAUSE CORONA

AND STREAMERS EFFECTIVELY EXTEND AIRCRAFT DIMENSIONS

STREAMERS

BASED ON GROUND MEASUREMENTS, AMPLITUDES LIMITED TO -100 A

U No MODEL BASED ON FIRST PRINCIPLES IS KNOWN TO EXIST FOR STREAMER

FORMATION AND PROPAGATION

S AT PRESENT, MODELING IS PROBABLY BEST ACCOMPLISHED BY PRESCRIBING

NONLINEAR CURRENT SOURCE REPRESENTATIONS

80

AN ANALYSIS METHOD FOR THE F-106 DIRECT STRIKE DATA

by

Dr. T. F. Trost

Texas Tech University

Summary of characteristics of electric and magnetic fields

and currents measured during strikes. Description of

laboratory modeling of direct strike fields and currentsincluding test apparatus, airplane model, and data acquisi-tion system. Comparison of model results with in-flight data:

Resonances, attachment points, and waveforms. First orderinterpretation of in-flight data as a lightning input-aircraft

response problem.

4

81

&- ... z

F-106 DIRECT STRIKE DATA

OBJECTIVES

I. MEASURE STRENGTHS AMD WAVEFORMS OF ELECTRIC AND MAGNETICFIELDS ON AIRCRAFT

II. INTERPRET WAVEFORMSA. DETERMINE WHICH CHARACTERISTICS OF WAVEFORMS ARE

DUE TO ELECTROMAGNETIC MODES OF AIRCRAFT/CHANNELB. INFER NATURE OF IONIZATION PROCESSESC. STATE IMPLICATIONS OF MEASURED WAVEFORMS REGARDING

COUPLING TO INTERIOR

APPARATUS FORAIRCRAFT-LIGHTNING MODEUNG

F - 1/2 IN. HELIAX CABLE

.!41 IN.SEMIRIGIDCABLE

36 IN. 1 8DOF- 106S-OMODEL SENSORS

/0- DOTJ

OSCILLOSCOPEWITH

CAMERA

PL.ANE

VERTICAL2 SAMPLING

INTEGRATORY'

62

0-DOT AT NOSE

1E-3 V

lee

-189

9 1.24 3.485.12 M 6 2.

IE-2 YS I-20

00

-20-3

-Q6

5.12 15.36 25.6IE-9 S

B-DOT AT FUSELAGE

IE-3 V

6848

-20

-Wl

-Ml

0 19.24 20.485.12 15.36 25.6

IE-9 SIE-12 VS

is

-25

-30

5.12 15.36 25.6IE-9 S

83

TRANSFER FUNCTIONSFOR THE F-106 MODEL

FREQUENCY DOMAIN DESCRIPTION

TRANSFER FOURIER TRANSFORM OF OUTPUTFUNCTION FOURIER TRANSFORM OF INPUT

INPUT: B-DOT NEAR LOWER WIREOUTPUTS: D-DOT AT NOSE

B-DOT ON FUSELAGE OVER RIGHT WINGB-DOT AT VARIOUS LOCATIONS NEAR SURFACED-DOT AT VARIOUS LOCATIONS NEAR SURFACE

EXAMPLES OF TRAN4SFER FUNCTIONS

1E-3 C4WD-DOT AT NOSE

380

2mIN

INso

a 19.95 39.89

9.97 29.92 49.8?IE6H

IE-3

B-DOT ON FUSELAGE

" 169ISO

40

1 I9.95 39.899.97 29.9 49.8?

IE 6 HZ

84

PRONY ANALYSIS OF FIELDSMEASURED ON THE MODEL

REPRESENT SENSOR OUTPUT WAVEFORMS AS FOLLOWS:N G-t

V(t) = I A e cos(wit +*.j~w1 1

= 2NR.e 1

s.i = cy.+jw.i ARE NATURAL FREQUENCIES OR POLES

R. ARE RESIDUES

FIVE LOWEST 1NATURAL FREQUENCIES "

OF F-106 MODEL

0 0

0

4c L

-4 7r

852

50 -- 1

40-

in 30.

Height ICloud- to- GroundTlightning Polk

Image Intro -Cloud

Charges lightning path

Model. of thunderStorni charge distribution.

CHARACTERISTIC TLMES

CORONA CURRENT RISETIME (LABORATORY MEAS.) 10-50 ns

LIGHTNING CURRENT RISETIME (REMOTE ELECTRICFIELD MEAS.) 30-8000 ns

PERIOD OF LOWEST ELECTROMAGNETIC MODE OFF-106 (LABORATORY SCALE MODEL flEAS.) 110 no

TIME CONSTANT FOR CHARGING F-106 IN*CHANNEL (ESTIMATE) -250 no

4 -106 DIRECT STRIKE RISETIME OF B 20-40 no

OF D 20-1000 na

86

T/S 1C-6 T

4W B9O 46 S-DOT

36

-498 15

-699 5

-sea aa .512 1041.536284 2.56 19.08 2.639.23 4.158.39

IE-6 S 1( 6 HZ1E-6 T

5 7 B

-16

-15

-2...........1.-4 2.4

.512 1.536 2.56IE-6 S

C/M2/S IE-6 C/112

3 0-DO

2.50

2

-5 1.5-19

-15 .

8, 1.024 2.048 a 29.16 48.31.512 1.536 2.56 19.98 393.23 59.39

IE-6 S 1E(6 HZIE-6 C/M2

* .5D

9

-5

-1 .5

-2.5

.512 1.536 2.56IE-6 S

87

CI25D-DOT 1E-6 C/4Q

28 80-038-05is 4

to 3 V101xi

* 2

-5

-1 I-- I. .4 2."4 a 3.16 4.31.512 1.536 2.56 19.18Ue 3 90.3

IE-6 S IE 6 NZIE-6 C/M

4

3.53D

21.5

a2.5

IE-6 S

T/S lE-6 T B-DOT

495 80-038-05 16

No 12

a a

- 4-40

.52 IE-6 S 2.5 18.98

IE-6 T8

.1 4

-4

.512 1.536 2.56IE-6 S

88

lE 6 A/S AI-DOT14

see 80-038-0514

3m8 lee I-DOT

18o 66

8 46

-100 28

a 1.02 2.04 a 4.13 81.25.51 1.53 2.55 2.06 6.19 10.31

IE-6 S IE 6 HZ

8

40

29

.51 1.53 2.55IE-6 S

SOME RESULTS FROM COMPARISONOF MODEL AND IN-FLIGHT DATA

IN-FLIGHT FREQUENCY SPECTRUM PEAKS ARE INGENERAL AGREEMENT WITH MODES OF MODEL

AT 9MHz AND 21 MHz

SHOULD BE ABLE TO INFER SOME CHANNEL PROPERTIESFROM IN-FLIGHT SPECTRA

rN-FLIGHT D-DOT WAVEFORMS LONGER DURATION THANB-DOT WHEREAS DURATIONS SAME ON MODEL

89

AN OVERVIEW OF THE ELECTRICAL/ELECTROMAGNETIC IMPACT OFADVANCED COMPOSITE MATERIALS ON AIRCRAFT DESIGN

by

Dr. John C. Corbin, Jr.

Wright-Patterson Air Force Base

The impact of the application of composite structures inaircraft presented from the electromagnetic viewpoint.Fundamental electromagnetic differences between metal andcomposite aircraft are described and technology develop-ments in shielding effectiveness, joint and fuel systemdesign, and power system/equipment integration are reviewed.

91

MNZ1IW PAM BLAI.NT 1

CA)z

00

zz

z a:

C/)

-J 00w CC CC

z Z w/ 00_0 cc co) x c

0 w *w C U0z I-J a

P C.) <~ C/

z z

a I- 0

0 UWz w

0 WW

w 0) -

92

I-

w

LLLZ

W~ U

m 0 .C/)C /

o CCC

LU 93

w -.- l

~LLJH

~ 00

L--H < CLI~r LL 0n< ILU~ < -y -j

0 j 0 0

Z UZ

(.:) LL)

H HI H I z

>L~ <~ H<QE, mP- -H <A0

0 Di IL a.

2L <

LLJu~< 10 0 z

'I U I -

U- z

Ln < > w94

AD~17 FDRLAITO DIITAINTEeNCLCNE l-T /

COMPENDIUM OF LIONTNINO EFFECTS ON FUTURE AIRCRAFT ErLECTRONIC--ETC CU)

UNCLASSIFED OT/FAACT2/30 N

moEEEohIUEEEEEEmhssmhhhhhhhhEEEEEEmoEEohE,EEElimEoEEohhoI

W ~I--I- i~C/) Cl)

I.-p wW W <

0 cc

Wc uW zwzccc8 0Z?0- 0

L0 0 0L

oL CC Z OO.

O)Z 0- Z> 0 0

jr Z CI~1. L3 Q CCOO) M.0 j

LU -J> cc -jWW ZZ _o<C)-

CM Wz z~

- - - L

w zZ 0~

(4)

wI

'-U.

"a<cc-a

4 cc

"Jog o -CO) Z c M U

00w0IUz c0r tcc0C0)W _jU

-4Q 3 !'

z LZ OX 96

w

C/)0

_ w LJ <-jLU IL w-z

cc 0 (1)

LU L D :0Cl)

LL <c 00Chin CD

C) U.oMw 0CZ0cn I- 0

Lj~ (1)c 0 >- X CLU 4 :CL0 0~ >

w W t. Cl)

co(nZ <

p 0 CC -J w

Lijc!) 2 0 w >z a m I

97

AIRCRAFT APPLICATIONS

*YF-16 FORWARD FUSELAGE

* F/A 18 HORNET

* AV-8B V/STOL

* 767

98

99

vi cm

0UU

co

10

LLIICA

zC

101~

00

Uc 00 0

WW

00ILzz

..7 .....

LLg

d -z

102

LJ

c- L&

f5 C),

LA C e _

C,,0

LLU

00

41

103

----------

LUJ

LLJ

C-,

LUA

II

'.4b

104

w zww

I-J -J w/ L

Iww c c~0>M - CL) 0/ (M)

10o W WC/Wi

C. - / UC)

QE -O0 C/) C L )cL). ZC/ WUJ C/)cWQ.aIC/

W 0. Z w

C/) C) zW j U

u 000* .

IC~c C F-F105

z

z wU

H1 0 05

w U) C/-CZ > < -j im

(LI U) -jLL )> zzz

La.

LU) z wLU 0 H.

0I L U LL cc 0 U)< ()) LU H

zo CL U) > w Low L 0 (0 w)~

cc U~ LLLuL H- LZ w

106

00

z 0-1

LL

LU

LUJ

C.:lC/ LI < L

w cc

co

CO 0 W0U

-J I- I- -

LLII ILI- 0.0 .>

107

R www

W) LLU U

L c .U LCl) Wa:w LrU..WzCL LZ<.

UOL Z W-j2 u L

COO (n U@0-W

0 c W -F-P1080

ok,,

xl I L

klCo

'no

~wU.

10910

L) U.-U U

02ww

&) C- )

I-

C/) wjw 0 0

C)w zU) C/) ~ z LL

o~C/) < <0P nz crO0O0Z mf

U. LL ~LL J )CCCW LL z < <<

0 -k uo 0

0 LLWr0cr0jLQ

110

0 -J

0 0U 0

w 04 0. 0

LU0 1,-Z LU z L0I("U w zi

cc >-f (JOw W ZjW4w

cm. 0j0 cc U

cm, -J 0 w~ 0 -wJ

LU cc.m w Z dI)

4 * * Su4 U

ccn=0 00M jiAc

W A

cc 1c1c

z0 -d

> W zlzi 0 00 U)

U- 0z wL cc

ZQ) Z l)or z aw wi

'4,R w Z-W WO<

0L % 0C/) C60

o/ C/) L)u. 0 OOC/)wLIZ I) C/) < 0 u

WL LL LLJ 0

C/ W LC/)LL

0./ C 0LSL L 0 U. cc 0 0.... ml'

*) 00 0

C..f) > LL L.LC/) L U LL112L

LU

U.'J

0 wLo < CD 0LLS z < LL

COO w m z 0CID C> i

1 L.C w -J wLw~~ Zl Z L:cl.:o~ 0 D.

Z(/) d < D o z

t:,-J -j -i -J 0 z_OOw w w w w w ~I

Z Z Z CD

0 0

.4c

113

cc 0 0

2 w L

LL 0 0

C,) w %-t -Ucc IC C ZLU _ w

x /) CCLJ 0LL a.WIC)ML .- uwL / _ N cLUJZc C) Uae <

LU -I -L

UfJ z

0 LL 0 C)W0 00C/) *- C)c

114

0~Q IL

00

x

CC

w4 0

LLLc-

11

0-0

W1

- a

CL c;

> 4-

00

0 _ _0 O

C.f) Ut

c-z

w

LA.

C,

II

0 0

00 0T 00

116

0

-J CD

CL 0'i CL 0 Q.

0~~~~ 0 I

0..C a N P z- >o) e wo

Z- C) 0- zLLJ= ZZ-

COO),~ .. U>< < 2MU

-n F-) CC 00 ~

JJ W cZ 0 Ow 0-

C/ 0. w 0 C/)J Z % z cra.. (n / z -w -

- - z .

L) L) a. F-

*0 0[[ 0uu .UL

117

xSn

Ca.0

N

'-4tz.

0

9.-'

0

14

118

* a-* * ,-~ .4 .t.

C.,

00

o 00

z

w N

0 CULON

.j.

L0

LUw

C,

Ei) -- SU 1311 UP04

11

:c3

00

CL -

-w >z~1 -

OO0

>1>

6: ClC

ww

w

Z0

wmNo0

LI- _ _ _

_j 0

Q 0 0

120

-j0 -Jz

COO) z w-1I wL

-,I 0-LJ M

HL -. Z -j 0 LH

i b WW w z0 )U) U - :,r) C/)

LLJ - LLH

-m 0LwwJ wo000 0 H~ rZ

0J w L>-:> C0(:5-) 0 0

cc ~~0. LUJ C/) U Cm~ ~ mw 2 L )z C/

E C C/) Z2 HHn %-. wL Z -.. Z

cc 0LJ 0 p p-j )D-J H L_ - zi

L.U -'0>LL - ZcrZ Z 0.

cc LL0. H

*V w 0

121

EMI LEAKAGE TEST PANEL LIST

CONF IGURAT ION

IDENT. CONFIGURATION JOINT JOINT SFAL CONFIGUR~ATION

A G/E Tape Hat'l None None

BG/E Cloth Wone None

C Aluminum None None

1) G/E (fIn Plated) None Norio

G OlE-Aluminum Single Lap Form-In-PIlice (FIP) Seal

Shear

G/OE-Aluminum Single Lap Tin Plated-FIP Seal

Shear

JG/E-Aluminum Single Lap FIP Seal

Shear

K GlE-Aluminum Single Lap HIP Seal-Finger Stock

Shear

L G/E-G/E Single Lap Tin Plated

Shear

P4 G/E-G/E Double Lap Fay Seal

Shear

N Aluminum-Aluminum Double Lap Fay Seal

Shear

P G/E-Aluminum Single Lap FlrP Seal-Finger Stock

Shear

R G/E-G/E Single Lap Tin Plated-FIP Seal-Finger Stock

122

TYPICAL TEST CONFIGURATIONS

To Receiver

Test Probe

Test ,.Specimen -' - -

/l-- - -- Test Enclosure

To Signal Source• / Xmit Antenna

GP03-0153-8

123

E --. .. . .-J,

EMI TEST PANEL DATARELATIVE PERFORMANCE OF PANELS WITH SEAMS

Relative

Test Panel Description Seam Preparation Performance

E H

N Aluminum-Aluminum None Double Lap Shear (OLS) 100 97

L Graphite/Epoxy Cloth - Graphite/ Tin Plated. Sealed 84 84

Epoxy Cloth

P Aluminum - Graphite/Epoxy Cloth Flonding Strip, Sealed 69 79

R Graphite/Epoxy Cloth - Graphite/ Tin Plated, fonding Strip 64 78

Epoxy Cloth Sealed

G Aluminum - Graphite/Epoxy Cloth Sealed 58 99

H Aluminum - Graphite/Epoxy Cloth Tin Plated, Sealed 49 96

M Graphite/Epoxy Cloth - Graphite/ None (DLS) 50 4

Epoxy Cloth

K Graphite/Epoxy Cloth - Graphite/ Bonding Strip, Sealed 37 55

Epoxy Cloth

J Graphite/Epoxy Cloth - Graphite/ Sealed 10 68

Epoxy Cloth

C (4) Aluminum 99 99

0 (4) Tin Plated, Graphite/Epoxy Cloth 95 89

A (4) Graphite/Epoxy Tape 80 71

B (4) Graphite/Epoxy Cloth 70 79

NOTES:

I. All panels show relationship to best panel whose value is 100

2. E (electric field)

3. H (magnetic field)

1 i4. One piece panels (no seams)

5. Single lap sheer (SLS) unless otherwise Indicated

6. The seals are a formed-In-place rubber gasket used to prevent water leaks

124

wo0<

ow UL. UU

LI- F-I-i-N L0 z' 0 cc

0m LW ZI-u-I zLJ >~

:z 0/

0 0D 0

0/ o/o5j

wl~cc L WLCC I-

125

ADJACENT GRAPHITE/EPOXY AND ALUMINUMLIGHTNING ATTACHMENT CHARACTERISTICS

Ioli in -JI

2 in,

1-0- 0

2 in.

0- c

J11

1261 P~ .6

- r--z--~A ~ ~ ~ I GrphteEpx

U-

(V E0) 4-

Zi Cl) ( 0 C)

L ) L).- 3 +- +

4) 4) - -

0 0 0 z z C) - L L-4U- -

L I- L , L L,

0 Wn V) (n ( n L U)(n. C) -l CL CL C

L c) O :3 a0 0 0VC ( V (

0.0(I) a, 0 ) INC) a

(V (V (V ( (VP (

-1 <) C) -) 0)0 0) D0z E I I I I E I0 0 - 0 U LL

<0

wz0.0

4: C L

zz

d 0

£1127

L :f 0 0 go 0 O 0 0

m0

- - N - C14

06E N

LU 8 N -4

C.) -0

0zi -r

U. 4-

0 o oW ~ F (A -4

o Op~~V) I'I- -

-w

(/ D <f C f

1 co (/) C.D x L0-I w1

C.) - ~ C 128U

z cc

0 z W

M Z 4 W .- U -40e W 0 0 z z0~

X M i0 I- L)Z

0C cc 00 J C)w

N C.)0 U

(1) z U i W ~jW/)w 0 z Zo 0z

L-S z.- W c

L U 0~ - z w

04 0 W 0 LL~ 4 ~M.LLJ - U C

0. zZ C/ 00 4Z 4

C7 z o -I LLL 01

0 W) LLJ W 1 - 4 - jc

LL ) LU. 4 - MW .Z aW W

0.4 Z) 0 0- a-W -2

ow z z ) 0 z~ >~W~ 0 Q0 jJ1 C C (7)J J L

4 - i r ~ w 0 :13

cn in( >O ZL L 00 WL-u..

V0 >0 L 0 0L

129

0) -

00

CocCV 4.'4-

LL)

QI x E

0 a- 0

I~CL

C.Y CL~0

wLL.

LL

13

z0

00

Cl) 2

000

00

m~ o

tL C

00

00

I-.2

I-D CL

/ !1c\i m 0-

CC

E 7@

131

w

- -c (flu.00 w

X 0*

c -

00~

zz r-

:1~~ c5

40J

132

00

X0

~0

Eo

_ _ E

C)~ C)

0. IC0 a)

3:- 0 c00

0 00 L0

_ _ _ N

<0 U

0 0coz C4C

A::SP061A

E3

cnc

cc <

LUU.0.

FAL

ow

00

00

w

Zw

C,

w,

134

COO)LUJ

COO)

COP) u

C.1) ccE

U cnc

< cc

L 0

I iT*C #O ,

<35

- . .- --------- - - -;-cc

wIW

CO) 0 <LL-J W 0

W- z Z =-

LLWW Z0' W cLJLJ c/) a. )c )LU0 RT~ A-W w

- W L o0 - L

<00 <OLLI

(/)C w0 McolIzF < 00'- L

> > Z ~0- CC LWF-~ wZOzz

w )WM0Z=

-10 0w 0 0

=) p: D 036

ui : 0 c

I,, I, V7l'xj ~

1.U

uI.

.2 'x

~13

0coz

zuJ

IzLU

W.

LU~LU

zu

z

1 38

EMI TEST SETUP IN ANECHOIC CHAMBER(A) ANTENNA, (B) FIELD METER, (C) WING

139

wall.

CC N ZLzL CD w i

w <LJ0w 0 0/)W

CO) ~ 0 Z cl W C/) -0I- <-~ oO C/WL Z U0o z~ zWmLU0 z w Z 1 <C

I- 0 O - WLJC/) zcLi Zz z 10LU 0w< < Z 0 Z

r 0 LL <Z)w :D cc 0< w

LL.L 0I-

LU o ZO C0 1 .. 0 o « CC

Z u n < <

ZU U- 0

I- Ow0 0 =.< I

I- WLL <O < < wc0w0 cc N ~ >0

w 0 0- LLZ C

140

<- 0 CC I WW

.j LJLLC.!W

LA. 0 DLJ ~ WwLA. m- - x

a.I z

I- aZZ

Z010

0u..2 0O

0 0

ui 141

0 c0 --LL 3z WI-

w wi wlZ ~~ rrZ L

I-U _ LL

Mr- LLL Z-J/)c

cc OZLLJ U LL

C/) W LJZWL

ZOO 0 wcrLL

1 " L. 13 j WC..> 0 % 5 Ow ZO

LL. >~I N 0

*A a 0 M

< C.)142

0 LLJ0 1= Lu 4

CIRCUIT AND PROBE ATTACHMENT POINTS

H 9i Vi, tag, PNobw Loc

/X

772

4/1

G if* -p? -

A l,,

10,

.\N

rAll d-mlh~s in inhesGl.il

143

Clo

z < 0 . w

L)wa C) <

al-I w z CC

0 00

c~LLcI WLL. 1

c c C !L0 WC C Z oZ

C.) L 0 Cl I144

LLI

wwLUJ

o oC. C/)M -j

W c

C5 0 C/)I ccw C

ca0 C/)-C/) w

c J ccw oo ~ 0 C/)0 0 00L.C/) LL 0 o

(nn cc 0_0 nZ c z C) C/o

Z 0co W WL wo U0

cc- 0 L w 0oe W U. C U. W

V CW

000

145

__ zCOLL <

CC 0

00 - :DCCLU--

_ Wz

00 Qu0LWLL ) e - C< ~ ( L/ 0LL

0) W0c

C)z 0U

-JwJ zwwLO~ J OI < -

00 c,0 z S U

146 o

SIMPLIFIED BLOCK DIAGRAM OF LIGHTNING TEST SETUP

St I

32

I0

0 10 26 0 500Ttme pusec Tm i

'Not -,5 ill iis

ICurrent C.Boied JIn

*.N.Botd on

Coupons

LLU

cz 75i

wl L LL

u 0J- i z- CM

LLU CO wi 0 CZOz w-LJ

ooLC.). C.

I- =zj o

LU uJ LU)

mx

LLLU

z miLUL

148

0-j

0

Q. -J LL)

Cu) 0 C-

-JLC

00

I-n

zWC

z U"

00

C) v) -J

14

ui C)

o -r

ui U, c

>I 004

F- u0

z>

C.)

C,

LU

zcCL0

150

wz w W

o0 C/)~(

cC)

DWCJ C/)O~W 1 L (. 0 Z F

ILJ- c/ w 0 Z 0~LL V)C)0o c0

LOjLLS W CL' p w .(Z/ C/)- C-) Z/ 0 -COO < 0 <. W j-.-

U- 0

0 L0z CC /)( C)U0 L.

151

aZ

ccL I 0aca. a.

C-a. I. C

W-w

- a., x

3 0 .4

-J -i b mwzo 0o L.

z - 2 S L

o-o~ C - a ca t

14 z w

::aa - mS~usus 1

ft ft ow tv .

wM A~ W.0 aaW4L% -aj

-~ a152

0

0L 0w l

Cl Z 0 - C/

Cd. 0 C) z F > LL 0

>0- > CC > L

<LLZ L3 W0- 0 <wj0uEnW w.m Z uI-

* 0 0i 0

Ei (- L I < -

0 0)

(I) -J!) < O C14

CC C/ CC FQ(i C

C7) c cc CV)w U a_1- 04 U

03 0 cc c -n0ccLr

c -L W zL Du ZO z c < W- -I-o CL0 Z c

z~c _L <cL)Fc 0 z Z

cc~ C J Q) 'C) toZ3O 0< Z/ E oo

<o w

o ? C 0 r 0 -i LO- 0 cc0 LW 0z wojw L

_ _- O z W ccwmC) -

-0 Z 0cc coD U0 D .1eLL~( .J< W a. u 0

ic u..ccY ( W 0 0

u..Gr D<c Fc0 0 uccw. a-~~ z 01.-j cc OQ

LLJm ;; Y b~l < CI. <Z.0 o cv, C) Z rj?' 5~ LLCCC)cc (.C 4 a. cc

*r LLW, c~~~- c-c CC>'- o -L)~i LdL wC /) o L x

(j Z 0 IfWl oc) I- o zcli C/) o L/)i (DD0Q

W Z T- DF- cnZ *F-L -

Zc a 0 cc jDW C) 0 >- D 02< 0

03 6 u.z 0 c coW~m 0<

:3 uj Go 1U) : -1 7 ;a L C

F- dc- LLJ154

STRUCTURAL APPLICATION OF COMPOSITE MATERIALS ANDTHE DIRECT EFFECTS OF LIGHTNING STRIKES

by

Mr. William E. Howell

Langley Research CenterNational Aeronautics and Space Administration

The direct effects of lightning strikes on compositematerials and protection approaches will be described.Approaches to EMI shielding which maintain weight advan-tage and structural integrity over the life of the

airframe will be discussed.

155

Mon -.4

SCOPE OF PRESENTATION

0 ADVANCED COMPOSITE MATERIALS

o APPLICATIONS IN AIRCRAFT STRUCTURES

0 LANGLEY'S RESEARCH ON DIRECT EFFECTS OF

LIGHTNING ON COMPOSITE STRUCTURES

MATERIAL PROPERTIES

PROPERTIES

STRENGTH, MODULUS, DENSITY RESISTIVITY, COSTMATERIAL KSI X 103 KSI LB/IN, OHM-CM $/B

GRAPHITE-EPOXY 110-225 18 0.055 0,9-1.1X10-4 30

KEVLAR-EPOXY 75-270 11 0.050 * 15

GLASS-EPOXY 75-200 6 0,070 • 10

ALUMINUM 55 10 0.101 2.8x10"6 2

DIELECTRIC MATERIAL (NON-CONDUCTOR)

156

FILAMENt PRE-IMPREGNATED TAPE LAMINATED STRUCTURE

GLASS1BORON

GRAPHITEf

KEVLARJ

MATRIX FlBI

EPOXY 1AR)POLY IMI DE

ALUM INUM J

IURI

~ -. 15.7

_ASSsi.

GRAPHITE-EPOXY SPOILERS FOR 737 COMMERCIAL FLIGHT SERVICE

FLIGHT SERVICE EVALUATION OF PRD-491EPOXY PANELS ON LOCKHEED L-1011 AIRCRAFT

159

COMPOSITESECONDARYSTRUCTURES

BOEING 727 COMPOSITE ELEVATOR

DOUGLAS DC-10 COMPOSITE RUDDER LOCKHEED L-1011 COMPOSITE AILERON

DC-10 COMPOSIE RUDDER

160

Lji~

V

BOEING 767 COMPOSITE STRUCTURE APPLICATIONS

r HYBRID COMPOSITE (KEVLARIGRAPHITE) STABILIZER TIPS

GRAPHITE COMPOSITE RUDDRF IXED T. E. FIXtD T.tL.

PANELS SPOILERS PANELS

INBOARD ELEVATORS

\- SEAL PLATES

so WING TO BODY/ -" ' FA I R IN G

0 0OUTBOARD0 A ILERONS

/. MAIN LANDING LEADING EDGEZ= GEAR DOORS \ ACCESS DOORS

NOSE LANDING (BODY)

GEAR DOORS COWL COMPONENTS

LIGHTNING PROTECTION SYSTEM FOR 727 ELEVATORF4A7

DECORATIVE PAINT 61790.05 cm (0. 02 in.) 6061 ALUMINUM

FAY SURFACE DSEAL/ADHESIVE-FIBERGLASS DIELECTRIC (2 PLIES) TYPICAL 120

_________________ I

A-AA-A ~99.06 __

/AA (39. 00 in)

STATIC DISCHARGEFITTING (3 PLACES)

FAIRING F IBERGLASS DIELECTRIC(2 PLIES)(EXISTING FIBERGLASS)" ,.05 cm (0.02 in.) 6061 LIGHTNINGSTEEL BALANCE WEIGHT DIVERTER STRIP

FAIRING(EXISTING FIBERGLASS)

161

A.2

AIRCRAFT STRIKE ZONES

* :'ZONE2A

-ZONES 1S AND 21

GRAPHITE/EPOXY UPPER AFT RUDDER 5176

DOUGLAS DC-1aF IBERGLASS -EPOXY

TI P ASSEMBLY

RIBS REA R SPAR

FRONT SPARALUMINUM ALLOY

DRIVE FITTINGS-.(2 PLACES)

FI BERGLASS-EPOXY'AUIMALOLEADIN EDGECONDUCTIVE STRAP

I FIBERGLASS-EPOXYI TRAILING EDGE

ANGLES

ALUMIUM ALOY LGRAPHITE(RH OPPOSITE

HINGE FITTINGS/ CLOSING RIB OiEDFRCRTY(5 PLACES)

162

STATIC DISCHARGEMAST BASE

LIGHTNING EFFECTS ON COMPOSITE STRU.CTURES

VERTICAL FIN CAP

LENGT 8 6 INI.Il

HEIGH 19 IN

WEIGHT 31 LB

MATERIALS

OEVLAR 49 EPOXY F-106B RESEARCH AIRCRAFT

GRAPHITE EPOXY

163

EXAMPLE PROTECTION SYSTEMS

o FLAME SPRAYED ALUMINUM

o ALUMINIZED GLASS (THORSTRAND®)

o ALUMINUM STRIPS/TAPE

o METAL SCREEN OR WIRE MESH

CERTIFICATION CURRENT TEST WAVEFORM COMPONENTSFOR

EVALUATION OF DIRECT EFFECTS

COMPONENT A

COMPONENT B

COMPONENT DCURRENT

I ,I

TIME

164

F-106B COMPOSITE FIN TIPS

o KEVLAR-EPOXY WITH THORSTRAND®

o GRAPHITE/EPOXY WITHOUT PROTECTION

o GRAPHITE/EPOXY WITH FLAME SPRAYED ALUMINUM

o KEVLAR/EPOXY WITH FLAME SPRAYED ALUMINUM

165

F1 068B COM PU1SI1T E F IN T IP A S SM ..Y

166

EFFECT OF LIGHTNING

ON

KEVLAR/EPOXY WITH THORSTRAND

UNPAINTED PAINU

p 105 IA ip 6 kA

ACTION INTEGRAL 0 82 106A2 ACTION INTEGRAL -0,41 106 A2Is

GRITE TRESE NO

KEVIAR IU

HylIRB MHICAL UPR

167

6L. '

168

CONCLUDING REMARKS

o COMPOSITES EXHIBIT STRUCTURAL ADVANTAGES COMPARED TO METALS

o EXCELLENT IN-SERVICE PERFORMANCE AND MAINTENANCE EXPERIENCE HAVE BEEN

ACHIEVED WITH OVER 150 COMPOSITE COMPONENTS DURING 8 YEARS AND OVER

2 MILLION HOURS OF FLIGHT SERVICE WITH NO SIGNIFICANT DAMAGE FROM

LIGHTNING STRIKES

o EFFECTS OF LIGHTNING ON COMPOSITE MATERIALS AND LIGHTNING PROTECTION

SYSTEMS ARE BEING EVALUATED

o GROUND AND FLIGHT DATA BASE IS BEING DEVELOPED FOR THE ELECTRICALSAFETY OF AIRCRAFT

169

-. ~.-

LIGHTNING INTERACTION ANALYSIS

by

Dr. Karl S. Kunz

Kunz Associates, Incorporated

Description of lightning interaction analyses and howthey are performed including lightning model, aircraftgeometry, interior equipment and cable layout, and model-ing of interconnecting system. Sample direct strike

problem and (generic) expected results. Overview ofmathematical basis used in interaction analysis anddescription of finite difference method.

'-I

171 MONiJ M MAN 94 nU

* -- *--- BE

WHAT IS A LIGHTNING INTERACTION ANALYSIS -

A MATHEMATICAL TREATMENT OF THE PROPAGATION OF

ELECTROMAGNETIC ENERGY FROM A LIGHTNING BOLT TO

AN AIRCRAFT AND ON INTO SUSCEPTIBLE INTERIOR

EQUIPMENT

USES EXPERIMENTALLY DETERMINED CHARACTERISTICS

AS THE INPUTS OF MATHEMATICAL MODELS

PREDICTS RESPONSE LEVELS (VOLTAGE AND CURRENT) AT

THE INPUTS (PINS) OF SUSCEPTIBLE EQUIPMENT

ALLOWS UPSET/DAMAGE/FAILURE ASSESSMENTS TO BE MADE

.FOR THE SUSCEPTIBLE EQUIPMENT

WHY A LIGHTNING INTERACTION ANALYSIS AND NOT JUST EXPERIMENT -

EXPANDS ON' EXPERIMENT

- SOURCE VARIATIONS, AIRCRAFT MODIFICATIONS AND

,EQUIPMENT CAN BE ACCOMODATED

INCREASES UNDERSTANDING

- MODELING INCREASES UNDERSTANDING AS MODELS ARE

REFINED AND MADE MORE ACCURATE

* EFFECTIVE

- ANALYSIS CAN BE QUICKLY AND INEXPENSIVELY PER-

FORMED, WHILE ADEQUATE ACCURRACY IS MAINTAINED

172

HOW IS A LIGHTNING INTERACTION ANALYSIS PERFORMED -

DEFINE SOURCE (DIRECT STRIKE, RADIATED FIELDS, ETC$),

AIRCRAFT (707, 747, ETC.) AND POSSIBLY SUSCEPTIBLE

EQUIPMENT (RADAR, NAVIGATIONAL EQUIPMENTj ETC.) -

OBTAIN RELEVANT EXPERIMENTAL DATA, SUCH AS

- LIGHTNING TIME BEHAVIOR OR SPECTRUM

- AIRCRAFT GEOMETRY AND CONSTRUCTION

- INTERIOR EQUIPMENT AND CABLE LAYOUTS

- ELECTRICAL PARAMETERS (CABLE IZE, SHIELDING,IMPEDANCE, TERMINATIONS, ETC.)

- DEVICE UPSET/DAMAGE/FAILURE THRESHOLDS ANDMECHANISMS

MODEL THE INTERCONNECTED SYSTEM OF SOURCE, AIRCRAFT

AND EQUIPMENT

SELECT A MATHEMATICALLY TRACTABLE REALIZATION OF THE

MODEL

CHECK THE MATHEMATICAL MODEL FOR COMPLETENESS AND

VALIDITY (COMPARE PREDICTIONS WITH KNOWN EXPERIMENTAL

RESULTS)

PREDICT RESPONSES AT THE SELECTED SUSCEPTIBLE EQUIPMENT

, EXAMINE SENSITIVITY TO PARAMETER VARIATIONS (OPTIONAL

"SAFETY" MEASURE)

173

r'PIAL PROBLEM

RESPQHS

LIGHTN$ DEFAILS

100KA Ax

174

AIRCRAFT DETAILS

air conditioner, cbnlgtcabin Pressure,power system contros

circuit e,.aacer panels fuel pumps and gauges

instrument panel niepesrs

oil pressure,tchometer,fire warning,generator,solenoid valves,

Land controls

Typical elements and cabling associated with the aircraft

power system.

tailcone

main deck

prumouri uipfn bayesubbulkhead

Elem taryvolues ft teuircrft.y

crwm welwlcab75

e4 i.en ba

CABLE DETAILS

CA lE TYPES

(TRANSMISSION LINES)

6 COAXIAL ......................

a TRIAXIAL........................

0 11ULTIAXIAL ............................

• TWINAXIAL ....................

I OPEN LINES

" STRIPLINE --

" MICROSTRIP .-

" TWO-WIRE 0 6

" EDGESTRIP --

" EDGEWIRE 0-

" WIRE OVER GROUND

MATHEMATICAL REALIZATION -

GENERAL REQUIREMENTS -

SOURCE X EXTERIOR RESPONSE X PENETRATION X INTERIORRESPONSE X SHIELDING EFFECT = RESPONSE AT PIN

* EXTERIOR RESPONSE X PENETRATION X INTERIOR RESPONSE

MOST DIFFICULT PART

POSSIBLE APPROACHES (MAY BE COMBINED WITH TRANSMISSION

LINE THEORY, BETHE SMALL HOLE THEORY, MULTIPLE RUNS, ETC.)

- SIMPLE CANONICAL SHAPES

- MOM THIN WIRE TYPE CODES (EFIE)- PATCH CODES (MFIE)- FINITE DIFFERENCE TECHNIQUES

176

= , .I

CANOJNICAL SWAE

z

z-h

-Ie 1117

grud ln

Diag am e va bradcrsed inidet an elecrocla tpherid.clidr

177

WIRE MMUEL/BOR 'S

'Itll)

21.6 mi/it -Ur1~l24.5m , N

/ C

24.3m - i33

8.5 M

ATM

Stic (intesectigcylnder)_modlofheb1_airraftnth

wing-fowr an wig-wp configurations

9. hi

26

Tom

.m : 1*..-

Body-of-revolution model of the EC-135 aircraft. Current zones

are indicated with dotted lines.

178

mom

Solle COMPUZZauoual Apects of Thin-Wire Mlodefing

2q-L -H L

F ~ 7 j2/5L

After MILLER L M4ORTON [4..203

After RICHMOND 14M.2]

After POGGPO L MILLER t4.21

After DIAZ (4.21After THEILE[(5=

Repruewtative wire grid model strucIt

1 79

FINIITE DIFFERENCE F-111 MODEL

/

- /

FINITE DIFFERENCE METHOD -

BASED ON BRUTE FORCE TIME STEPPING OF MAXWELL'S

EQUATIONS

INCORPORATES ANY MATHEMATICALLY EXPRESSIBLE' SOURCE,

I.E. (t) = e t - e Bt

" ACCURATE EXTERIOR RESPONSE PREDICTIONS

* WITH EXPANSION TECHNIQUE CAN, ALSO, PERFORM INTERIORRESPONSE PREDICTIONS AT REASONABLE COST

CAN HANDLE COMPOSITES, AS WELL AS METAL AIRCRAFT

PANELS

180

.,-,m

DEFINITION OF THE FINITE DIFFERENCE APPROACH

Finite differencing consists of replacing continuous

partial derivatives in P.D.E.'s with appropriate finite

differences. For example:

y - A (yn) - Yn-l " Yn ; 'yA [A(Yn) Yn+2" 2yn+ + Yn ; etc.

Finite differencing discritizes the P.D.E. and, hence,

the problem being solved, i.e.

Finite differencing is, therefore, an approximation

that in the limit of zero mesh size is exact.

Finite differencing does not require any special model

of the problem to facilitate a solution, just the appro-

priate P.D.E. such as:

I- T V2T + Q/ k or ( r - " e" ) -"

(heat equation) (portion of Maxwell Equationsfor thin scatterer in cylin-dri cal coordinates)

181

EVOLUTION OF PRECEEDING FINITE DIFFERENCE APPROACH TO EM

t _ ., 4.-

• The Maxwell Equations

" Classical Boundary Value Solutions

• Introduction of Computers/Computer Oriented Numerical Analysis

" Integral Equation Approaches (EFIE and MFIE)

" Finite Difference as presently applied to EM coupling

1) feasibility of application to realistic problems -

K.S. Lee - "Num. Sol. of Int. Bound. Val. Prob. Involv-

ing Maxwell's Equas. in Iso. Media", IEEE Trans-

actions A and P, May,1966.

2) application in 2D to realistic problem with radiation

boundary condition -

D.E. Merewether - "Trans. Currents Induced on a Metallic

Body of Rev. by an EM Pulse", IEEE Trans-

actions on EMC, May, 1971.

3) formulation of 3D code with radiation boundary condition -

R. Holland - "THREDE: A Free-Field EMP Coupling and Scat-

tering Code", Mission Research Corp., AMRC-

R-95, Sept., 1976.

182

EVOLUTION - 2

4) application to complex scattering object with com-

parison to experiment -

K.S. Kunz and K.M. Lee - "A Three-Dim. Finite-Diff. Solu.

to the Ext. Response of an Aircraft

to a Complex Transient EM Environ-

ment: Part 1 -The Method and Its Im-

plementation and Part 2 - Comparison

of Predictions and Measurements",

IEEE Transactions on EMC, May, 1978.

5) expand subvolume in a second run for increased spatial

and frequency resolution -

K.S. Kunz and L.T. Simpson - "A Technique for Increasing the

Resolution of Finite-Difference

Solutions of the Maxwell Equations",

IEEE Trancact on on EMC

November, 19816) generalize the 3D code to treat lossy dielectrics -

R. Holland, L.T. Simpson and K.S. Kunz -

"Finite-Difference Analysis of EMP Coupling to Lossy

Dielectric Structures", IEEE Transactions on EMC,

August, 1980.

183

Law

33 ~ ~ ~ ~ t r EEZfWANS.ACT IONS~. ONELC a I C OMPATI BIkLT'V. V OL. Er.IC-20 NO.-~ A~y 07b

-. 0 2t0 500o 790 1_~0 1 0 250 500 750 --- ro

FiS. 17. T.P. R[L. jAf). EI Fuselals. Fig. 21. TP. 382.jA(IJ.EU IJ.5S.e

10 ...

2' U& 0U '

1Fig. 22. T.P. 3 10j aFm v

Fir-I& 382.JA, .).)I Fmebs.

~~Ii U~0.25__ _ _ _fir.. -0.2S-

0 250- Sao 80 1

o0 250 500 75 1000 Tire (ns)Fi&.23. 1.?. 332, Qf).£g Fushlqu

-Fig. 19. 7.?.RI.Q2(t.EU Fuw.Ia.

:\00.000.-

10 100

SAMPLE PERFECTLY CONDUCTING A/C RESULTS '

184

INCREASED RESOLUTION USING EXPANSION TECHNP*UE-A/C UNEXPANDED-

toc-. ,o ~ A~ ft 4m 1. so

N,.S00." 1it

-A/C EXPANDED -

PlgweO 7. Fitoended Amg Showinsg Cockbit Arma Dtail.

185

S

5

L I

I

'I r. ~-~ - -

~4:~ fI *-

.j.

I15-

S 4 ' 9

9, (9111,9 * *9

j.j*I)~)- * 3 *

-'

5'

I

E.if

I*

LIJ 1

I a -I

7 ,., a:

-I Si *~' I I

p J I.; Z i'. -

z Ii'

-

U* S

K * S

(.9.) ~* -J

Sn I

I x N, e

- I .3 I

I- 1 g~~qq~ -S

-J SM ' *

-~

I*SSM1S*,9U.

- U *

La..I

U, 5'

cd~ S.

-J

B31

a!'I

t

12

- .9

-, ii- up

* - aa* -. Lg

186

S.- -

7

S

*

6

*6 ..~

*

* I

I

.1 I..

_________ 1. I

I I:

__________ I

U I*

U U I

~1

(WhUl:~ - . . .

.

I (111W. 3 - -

1II

M p

SbLb

-

* S

Pd

P. -,

4 I I

& I

UU.*1

I I*

U.

U.

U. * I -

.1;.5,, * -~ U t.

.~ *t-~-~

I! U jJ~... .

-- II .5 *~*Y,,* -

-. ** OS

-£ I

ba a

*

.0

Lb

U.1.5

I I I

I I I U* I

I 5

1

4 .1

.2 U. U.

-*~I

U '5 U

-I U .1

-~

LI

I~IuI~2- * ,I ..a I-,3I*.Iu

* * * Sb

* 3lb

Pd

P .9 .9lb

187

*~' ~

IN1TERIOR COUPLING APPROACH

Contiguous Ski" "Ide of the Aircraft for Diffusion Calculatin

Perfectly Conducting Airtraft with Sem

Cmplement of the sow i'adel

ORIGINAL GEOMETRY CONJliGAr/ftONPLEM4ljl GtIOJRY

Electrically Conductingj I ~ow ,Screen SI lcrclI ulu I"

I S. ren.

"total

I>I

E lectrically /I.MI1eialIConducti ng 1:f i.onduc I I nq

Source sI Cwjat

flcric aly AprueMagnetically

(oI I Il petr Conduc tIg w)Wire

Origjinal and Counjuwateltnmnplenment Geoemtry-1, lanar Screen txample

188

Escat. -E nc af(t)tan tan/ I / Electrically

Conducting

EatResistive (iretan t. (E H 4s )R/L Segment ( emns

tnloop ofRes istance

R

Escat. -E inctan tan *

___________________Aircraft Surface

Lightning Channel Source

Hscat. -H mc ug(t)tan tan

Magnetically(E E1as1) onjugate Conducting

,sa.lo Resistive Wiretan RL Segment (3 segments)

Hscatl Hinc 0tan Htan

Comnplementary Aircraft Surface

Conjugate Lightning Channel Source

189

I 1 tM~ ~ II So' Id. nAI% t

Orgia Sufc nSt[ A -AE, f

no zeroed

I7drcl I

wire

finit Difrnefruaino/h CmlmnayArrf ufc

Giin t he ufer ce rwauoofteCmpetayAcaf Srfc

AttacimientChannel

.uruace tal taH 0

Resistiveen ChannelnCwphereuae/tnpeni C-- ea tr as It rs laO ifr Fnt tfee c psi;I

(YE190

,,sea, I "

PIN PREDICTIONS - SINGLE MULTILAYERED CABLE

m-

\ GNU

............ f hu~a.......

Moee nf aerW cableLModel of * cabim

USE F.D, TO FIND CURRENTS ALONG CABLE SHEATH,Ib

USE SURFACE TRANSFER IMPEDANCE, ZT AS DEFINED BY

SCHELKUNOFF, TO FIND THE INTERIOR CURRENTS: -*Ia(x) =J G1 (x,x') -ZT , Ib(x')dx'

WHERE Gil IS THE APPROPRIATE GREEN'S FUNCTION THAT

INCORPORATES THE CORRECT BOUNDARY CONDITIONS FOR

THE GEOMETRY AT HAND

EVALUATE la (x = L) TO FIND THE. CURRENT AT THE PIN

EVALUATE THE VOLTAGE ACROSS THE DEVICE CONNECTED TO THEPIN, Vd, USING THE PIN CURRENT AND THE DEVICE'S IMPEDENCE

Zd, I.E. Vd = Ia(X = L)Zd.

ONLY DIFFUSION THROUGH SHIELD CONSIDERED

191

A7 A114 117 FEDERAL AVIATION ADMINISTRATION TECHNICAL CENTER ATL-ETC F/6 1/3

COMPENDIUM OF LIGHTNING EFFECTS ON FUTURE AIRCRAFT ELECTRONIC-(TC(UI

UNCLASSIFIED DOT/FAA/CT-62/30I ~ EnnunEnunuuIlEEllEEllllEEE*EEEIIIIII

EXPECTED RESULTS-

* EXTERIOR RESPONSE -SIMILAR TO PERFECTLY CONDUCTINGA/C RESPONSES

*INTERIOR CABLE RESPONSE, WIRE EXTERIOR - WEAKLYDAMPED SINUSOID, SEE BELOW

PIN RESPONSE -SIMILAR TO WIRE EXTERIOR

Interior Geometry of Exoended Recion

10's .

ow i: __ _ _i.0 .es 10' :111110 100 200 300 400 107

T i m e ( n s )F r q . ( H Z )

Smol of reditedResults

192

SUMMARY

FEATURES

o Any fuselage exterior geometry can be modeled, including:

- composites

- conducting panels

- mix of composite and metal

o Various coupling pathways can be modeled, including:

- diffusion through composite panels

- seams around panels

- small apertures around doors/hatches/wheel wells, etc.

- large apertures such as windows

o Detailed interior geometries can be treated, including:

- individual wires with various terminations

- surrounding "fill"

- adjacent wires to within a cell size (-O.05m or -2" at best)

o Fast risetime (~30-5Ons) pulses can be easily incorporated

o Non-linear effects can be modeled

- attachment points can be selected based on experience, whiledetachment points can be selected based on fields exceedingpreset thresholds

- interior arcing can be modeled similarly

193

AREAS OF APPLICATION

0 Exterior I and V response predictions as a function of:

- position

- A C construction (metal or composite or mix)

- excitation source

- attachment/detachment location

o Interior Responses

- interior field levels

- wire currents

- transfer functions

o Hazard Assessment

- induced current damage

- field induced upset

- fuel ignition from arcing

o Protection Measures Evaluation

- field/charge and current penetration reduction(from covered seams, mesh across windows, etc.)

- arc suppression(from cable rerouting, interior geometry changes, etc.)

- protective device effects

1 (depends on "threat" spectrum, device location, deviceoperations, etc.)

INTERMITTENT/TRANSIENT FAULTS IN DIGITAL COMPUTERS

by

Dr. Gerald M. Masson

Johns Hopkins University

Need and objectives of digital system upset assessmentmethodology: The definition of upset. Description ofapproach being developed for assessing upset potentialof digital systems. Upset model types and importance ofburst error models; dominant importance of program transi-tion models. Definition and use of system state probabilitytransition matrix in upset analyses. Example of upsettolerant microcontroller.

195

EAULL ANALYSIS EAUL TOULERANC

* FAULTS * DUPLEX SYSTEMS

- PERMANENT - ROLLBACK

- INTERMITTENT * TMR SYSTEMS- TRANSIENT - ROLLAEAD

- I/T * DIAGNOSABLE SYSTEMS

* MODELS - INTELLIGENT UNITS

- STUCK-AT

* TESTS- DETECTION

- LOCATION

MICROPROCESSOR CONTROLLERS FAULTS IN MICROPROCESSORS

HARDWARE * CPU TESTING- CENTRAL PROCESSING UNIT (CPU) ' RECOVERY STRATEGIES

- READ-ONLY MEMORY (ROM) - DUPLICATION

- READ/WRITE MEMORY (RAM) - TMR

- INPUT/OUTPUT DEVICES (I/O) - WATCHDOG TIMER

- SUPPORT LOGIC

* SOFTWARE- LOOKUP TABLES- CHARACTERIZATION

I. INPUT SENSORS SCANNEDI. DATA PROCESSINGIII. CONTROL SIGNALS TO ACTUATORS

196

i

IWATCHIDOG TIMER

M ICROPROCESSOR

RESET T2T

INPUT

INPUT

x

RESET

INPUT

FIGURE 1. WATCHDOG TIMlER

197

I--

DAV~

HL

19

LIC.7

OPERATING INPUT CONTROL OUTPUT

ENVIRONMENT SYSTEM

!CONTROL

FIGURE 3. CONTROL SYSTEM SITUATED IN ITS ENVIRONMENT

EQUIVALENT

~FAULT

I'-- GENERATOR

FUNCTION

INPUT TRANSFORMP ER TURBATION INPUT

OPERATING INPUT EQUIVALENT OUTPT

ENVIRONMENT CONTROL SYSTEM

CONT ROL

FIGURE 4. CONTROL SYSTEM SITUATED IN EQUIVALENT HOSTILE ENVIRONMENT

199

FAULT/SYSTEM I NTERACTI ON

* CONTROL SYSTEM INTERACTS WITH ENVIRONMENT

* I NTERNI TTENT/TRANS IENT FAULTS- SYSTEM OTRANSIENT M RESPONSE- SYSTEM STEADY STATE* RESPONSE

[FAULT SOURCES

n FUN:TC)NNFPLIT RNF M

FAULT

LEVEL

FIGUR NM5. CONTROL SYSTEM STAE NHSL NIOMN

200

FAULTS. ERRORS. AND UPSETS

* FAILURES

- CIRCUIT LEVEL

- ANALOG NATURE

* FAULTS

- LOGICAL LEVEL

- DIGITAL NATURE

- LOGICAL DIFFERENCE AT FAULT SITE

F ERRORS- LOGICAL LEVEL

- DIGITAL NATURE

- DUE TO PROPAGATION OF FAULTS

- LOSS OF SYNCHRONIZATION PROBLEM

NEW DEFINITION

* UPSETS

- SYSTEM LEVEL

- FUNCTIONAL NATURE

- TRANSFER FUNCTION

CONTAINMENT SETS

* A FINITE SET OF MUTUALLY EXCLUSIVE FUNCTIONAL STATES

* COVERS ALL POSSIBLE SYSTEM TRANSFER FUNCTIONS

* INCLUDES

- VALID STATES- ERRONEOUS STATES

201

.=-. . ... ; :... .I .. . ./ --° - ' qb - . - , ' '" :..,,.T_ '. ,,,=. '

4-

CONTAINMENT SET TRANSITIONS

TRANSITION MATRIX:

T-1il ,WHERE Pjj IS THE PROBABILITY OF A TRANSITIONFROM STATE i TO STATE i ,GIVEN THAT AN ERROR HAS OCCURRED

[Po]

L (K+l) = T L (K) , L (K) PIiI

Pn'

p1 (K) IS THE PROBABILITY OF BEING IN CONTAINMENT STATELi £ (L) AFTER K UPSETS

AFTER K UPSETS,

L (K) = T KL (0)

FAULT ANALYSIS AND SYSTEM VALIDATION

*SINGLE UPSET ENVIRONMENT

-P 1 1 OF T[lP.] OF MAJOR CONCERN

*MULTIPLE UPSET ENVIRONMENT

-P 11OF TK FOR LARGE K OF MAJOR CONCERN

*EXAMPLE

= 15/16 T~= 78 1/8111/4 1/16 ,18 7/8

[1 T*K5/19 15/19] I (*) [1/2 1/2]

14/19 4/19 , - 1/2 1/2J

202

1/4

5/16

B) T*1/8

7/18

1/8

FIGURE 6. TWO 2-LOOP IMPLEMENTATIONS

FAUL TOLERANE

° CONTAINMENT SET AND TRANSITION MATRIX ANALYSIS PROVIDES FEEDBACKTO IMPROVE DESIGNS

" REMOVAL OF ERRONEOUS ELEMENTS FROM THE CONTAINMENT SET

AUTOMATICALLY PROVIDES FAULT TOLERANCE

203

CONTAINMENT SET FOR MICROPROCESSORS

° PROGRAM TYPES

- EXITING

- LOOPING

" TAKE THE CONTAINMENT SET TO CONSIST OF ALL POSSIBLE LOOP PROGRAMS

- INCLUDES VALID LOOPS

- INCLUDES ERRONEOUS LOOPS

GIVEN THE EXECUTION OF LOOP PROGRAM Li c {L) UPSETS CAN

BE CHARACTERIZED AS:

(1) DATA CHANGE --

DATA VALUES ARE MODIFIED, BUT EXECUTIONREMAINS IN Li

(2) PROGRAM BUMP --

EXECUTION TEMPORARILY DIVERGES FROM Li BUT

EVENTUALLY RETURNS TO Li

(3) PROGRAM TRANSITION --

EXECUTION JUMPS FROM L i TO L.A , L L

PROGRAM TRANSITION = STEADY STATE OPERATIONAL DEVIATIONTRANSITION INTO INVALID EMBEDDED LOOP = SYSTEM CRASH

204

hflRECONTENTS

NONtIAL EXECUTION N OP-CODE

ERRONEOUS EXECUTION N+1 DATA

NORMIAL EXECUTION N OP- CODEERRONEOUS EXECUTION N+1 DATA 0ERRONEOUS EXECUTION N+2 DMIA I

ERRONEOUS EXECUTION N DATA TABLEERRONEOUS EXECUTION N+1 DATA TABLE

FIGURE 7. ERRONEOUS LOOP INSTRUICTION EXECUTION

R~esult Addr Data LO Code Li Code L2 Code other

LO 0000 00 NOP

L; 00C9 00 OLO 0OCA 33 X->VI AOC3RLi WEd C3 -JMP 0D3HLO oocc D3 OUT 0Lo OOCD 00 N OPLO 00CR 22 LXI ROCFC3HL2 Od? C3 C -~3 PRLo GODO cvp I IMTLO ODI 00 POP IILO 00D2 32 STA OCDC3H SLi 00D3 C3 imp OdIN

4Lo GOD4 Ca RITYLo GODS 00 NOPLO 00D6 C3 imp OCABLO 00D7 CA 33 0LO GODS 00 POP

Lo 0009 00 POP

Lo 001 00uoLO 001? C7 RE? 0

Figure U. An Erroneous Loop Example for the 305

205

DATA BU .

MICROPROCESSOR ROM RAM 0

SELECT SELECTrLC

ADR5 ECODE DECODE DCOE

__ BUS

PROGRAM-PATH DJVERTERS

*HARDWARE *SOFTWARE

- CLOCK STOP - JUMP CATEGORY INSTRUCTIONS- HALT - CALL CATEGORY INSTRUCTIONS- READY - RETURN CATEGORY INSTRUCTIONS- HOLD (+ EFFECTS OF OUNDEFINEDv

2 - INTERRUPTS I NSTRUCTIONS)

SA PROGRAM STRUCTURAL ANALYSIS IS MADE BASED ON PATH DIVERTERS

206

9O4---

ULT T N O E DSI CALLE

-OTINOU w LC up

UN R 0 LouT OP ISTRUCTU EXS

(OP-CODE FETCH DETECTION)

u-i- - BAD FETCH DETE.CTORIL q'(ROM RESTRICT MECHANISM)

- OPTIONAL SAFE ROM

- REQUIRES 2-3 INTEGRATED CIRCUITS

207

- COTNUU CLC

-UNONDIIONL INTRUCIONEXECTIO(O-CDEFEC DTETIN

TOALE I

CRASH-PROOF SOFTWARE DESIGN RULES SUMARY

1. PROGRAM EXECUTION FROM RAM Is PROHIBITED.

2. EXIT FROM TEMPQRARY LOOPS MUST BE GUARANTEED.

3. STABLE LOOPS MUST INCLUDE A CALL TO A CHECK SUBROUTINE, WHICH GUARANTEES ALL

ASSUMPTIONS THE PROGRAM REQUIRES FOR CONTINUED EXECUTION. (THESE INCLUDE

ASSUMPTIONS ON 1/0, INTERRUPTS, AND VARIABLES.)

L. RETURN FROM ALL POSSIBLE INTERRUPTS MUST BE GUARANTEED.

5. SUBROUTINES CANNOT CALL THEMSELVES.

6. SUBROUTINES CANNOT MODIFY THE STACK POINTER OR THE RETURN ADDRESS.

7. WITHIN SUBROUTINES, MEMORY STORE INSTRUCTIONS WHICH PERMIT A VARIABLE STORE

ADDRESS MUST GUARANTEE THAT THE REGISTER USED AS THE ADDRESS POINTER CANNOT

POINT TO THE STACK SPACE.

8. INSTEAD OF USING RET OR RCR INSTRUCTIONS, A JIP OR JCN TO A SPECIAL RETURN

ROUTINE WHICH GUARANTEES THAT THE STACK POINTER POINTS WITHIN THE STACK SPACE

BEFORE RETURNING MUST BE USED. (THE RETURN ROUTINE CONTAINS THE ONLY RETINSTRUCTION IN THE ENTIRE PROGRAM.)

9. A STACK WALL MUST BE ADDED. (LEAVE UNUSED RIM SPACE AS 00 OR FF.)

10. THE PCHL INSTRUCTION CANNOT BE USED.

, 11. EITHER A SAFE ROMI MUST BE USED, OR ALL ERRONEOUS LOOPS AND ERRONEOUS CALLS

TO ADDRESSES WHICH, WHEN CONSIDERED TO BE SUBROUTINES, DO NOT SATISFY EITHER

RULE 5, 6, 7, OR 8 MUST BE REMOVED.

208

FAULT TOLERANT CONTROLLER TEST SYSTEM

" NOISY POWER SUPPLY TEST SOURCE

" DUAL LED CONTROLLERS

* UNMODIFIED SOFTWARE

- LED 1.1

- LED 4.6

° CRASH-PROOF SOFTWARE

- LED 2.2

- LED 3,3

- LED 3.4

- LED 5.2

- OVERHEAD

LED 2.2/LED 1.1 = 8.5 %LED 5.2/LED 4.6 = 14 Z

LED 3.3/LED 2.2 = 35 ZLED 3.3/LED 1.1 = 47 Z

SOFTWARE TOOLS

* DESIGN AIDS FOR FAULT TOLERANCE IMPLEMENTATION

" INTERACTIVE USE PROVIDES EFFECTIVE AND EFFICIENT DESIGN

- SAFE

- PRODUCES SAFE ROM CONTENTS FROM SOURCE CODE

S, - OUTPUT USED AS INPUT TO LOOP

LOOP

- LOCATES BANNED PROGRAM STRUCTURES

- LOCATES ERRONEOUS LOOPS

- PROVIDES CHECK ON INTENTIONAL LOOPS

209

LOOP ANALYSIS OP LED 1.1, FAGE I

P LE . .OBJLED.. 1 .*SAF'

ONLY VALID (Vv) OR ERRONEOUS CC*§') ? ELOOP: ERRONEOUSO0IBrD PSH PSWGOFC 00 HOPOPDg C3 r3 O0 JMP OFscar3 05 D(CR 800r4 CS RZ

LOP SEAEEH? NLST CALI O? YONLY VALID CVv) OR RRONEOUS CC,e') ELOOP: ERRONEOUS

036D PC 07 rB Cm r907 [1/0

LOOP: ERRONEOUS0395 rc OP PC CM FEOr [1/0]

LOOP SEARCH? NLIST CALLS? NLIST MEMORY STORES? YONLY VALID (V,v) OR ERRONEOUS (Et') ?VLOOP: VALIDDD6A 77 NOV MA

LOOP: VALIDOOAC 34 IN M

LOOP: VALIDDon3 36 o0 NVI M, 00

** MAIN ANID SWTCH SET DS ERRONEOUS **

% LOOP LEDI.I.OBJ LCDI.i.sArcLOOP SEARCH? YONLY VALID CV~v) OR ERRONEOUS (E.,r) ? VLOOP: VALIDO09D CD EB 01 CALL 0CE4OOAD OD DCR COOAI C? 9D 00 JNZ 009D

LOOP: VALID0OP3. 05 iCR I:OOD4 CS RZO'S Fa INX H006 7E MOV A*MOF7 P3 INX HOars F rO CPI PO

4 ODFA C2 F'5 O0 JN7 OOr5OarD C3 P3 00 JMP 00P3

LOOP: VALIDOOF5 23 INX HO0P6 7E MOV A,.OOr7 23 INX HOO' PC F'O CPI '0GOFA C2 F5 00 JNZ Or5

LOOP: VALID00F3 05 DCR 800P4 CS RZ0QF5 23 INX HOP6 7E MOY AM

210

im - " .84 I M40a ".4 4oC £090 1 S .a 0 060.0 0on 0 us.

* .4 '.1 toI *4 . 4

54 C Iwo 0UMJ. 04'

4i 0 0 49 041 0 405

0

.45 4

N0 inP an " 4

944OA

-21

C)

LLLU

C,,

C/) I--

F- W-

Lo 2

Pu-l C/)lLoLo9 CL)

F-1- CLJ Z .,

F- F- CI- - 0-

C)Cl D-" Li C)4lCe) Eb-" l M - C/)

C: LDL J - - LI CD /)0 L C) _j = C-) ) 0 5

C- LL C/l)C F- V) MC0) -2

I LLI I;:A -

212~

A MICROPROCESSOR-BASED UPSET TEST METHOD

by

Ms. Celeste M. Belcastro

Langley Research CenterNational Aeronautics and Space Administration

Description of a microprocessor-based upset testing

method employing transient waveforms randomly injectedinto a digital unit. Upset test data presented along

with preliminary observations.

213

OUTLINE

9 BACKGROUND

0 RESEARCH OBJECTIVES OF UPSET TESTING

S'UPSET TEST DESIGN CRITERIA

0 UPSET TEST HARDWARE IMPLEMENTATION

0 PRELIMINARY OBSERVATIONS

0 FUTURE PLANS

~t

o 0!

214

INDUICED FFFECTS.TESTING LEVELS

*SYSTEM AND SUBSYSTEM ASSESSMENT.-

CABLE EXCITATION

*CHANGING MAGNETIC FIELD IN A COUPLING TRANSFORMER

*TRANSVERSE ELECTROMAGNETIC WAVES GENERATED ONPARALLEL-PLATE TRANSMISS ION LINES

* INDIVIDUAL UNIT ASSESSMENT-

INTERFACE CIRCUIT INJECTION

*DIRECT APPLICATIOI OF TRANSIENT WAVEFORMS TO PIN CONNECTIONS,

RESEARCH OBJECTIVES

6 SHORT RANGE

*DEVELOP A METHODOLOGY TO TEST A DIGITAL

SYSTEM FOR UPSETS

.0 LONG RANGE

I CHARACTERIZE UPSET PHENOMENA

*DEVELOP DIGITAL FAULT SIGNATURES THAT

MODEL UPSETS

- DIAGNOSTIC EMULATION

- UPSET VULNERABILITY ASSESSMENT

215

UPSET TEST CRITERIA

* CHOOSE A CANDIDATE PROCESSOR/CONTROLLER

0 INJECT TRANSIENT SIGNALS RANDOMLY

0. ENHANCE SIMULATION

* AVOID SYNCHRONIZATION

0 IDENTIFY PROCESSOR'S INTERNAL STATE WHEN INJECTION OCCURS

*'DETERMINE IF UPSET .OCCURS INDEPENDENTLY

OF PROCESSING STATE

0 VARY THE TRANSIENT SIGNAL INJECTION POINT

* DETERMINE IF UPSET SUSCEPTIBILITY

IS UNIFORM THROUGHOUT THE

PROCESSING SYSTEM

0 OBTAIN BIT PATTERNS

* DEVELOP FAULT SIGNATURES

CANDIDATE PROCESSOR/CONTROLLER

INTEL pP UNIT-

ADDRESS BUS

DAABSOUTPUT krDAABS u DATA

COK0, 8080 F MEM~ORYF

CKTS, 0. CPU 6E-- R

INPUTT STATUSINPUTA WORDREADY mux TLINE.U

C014TROLFROM CPU

INPUT' DATA

216

PROGRAMMING-HI ERARCHY

PROGRAMMER

8080 INSTRUCTION SET(24i4 INSTRUCTIONS)

PROGRAM

OPERATION CODE (OP-CODE)

8080 PP

INSTRUCT ION CYCLES

MACHINE CYCLES (IDENTIFIED BY STATUS BITS)

STATES

CLOCK CYCLES (02)

MICROPROCESSOR MACHIINE CYCLES

TVPUE MACHINiE SIAIUJIMl M0JQL CLOCK--CYCLESCCLE S/i S. sa s! Sa 52 SI so

INSTRUCTION FETCH 1 0 1 0 0 0 1. 0 4~ (11MFMORY READ 1 0' 0 0 0 0 1 0 3MEMORY WRITE 0 0 0 0 0 0 0 0301%4STACK READ 1 0 0 0 0 1 1 0 3STACK WRITE 0 o a 0 0 1 0 0 3

jjINPUT 0 1 00 0 0 10 3

OUJir 0 0U 1 0 .0 u 03INIERRUPT 0 0 1 0 0 0 1 1 .5HALT 1 0 0 0 1 0 1 0 3xINTERRUPT WilI LE HALT 0 0 1 0 1 0 1 1 5

217

UPSET TEST CONFIGURATIQN

I TRASIENT IANSIENT f DIGITAL UNIIT

INJECTION sI C~. N - UNDER l~EOTlL SOUR1CE RA

COAROL LO C WE pP 11

TIME

TIMR I UPSET ERR

I DETECTOR ~

SI UNPERTURBEDTRANSIENT SIGNAL REIlENCE ___

INJECTOR/ECORDER ON I I"RA

imrEL Itp 02

~~ UPSET flFiECTOR/RECORDEll

TELETYPE 1/O

LIGHTNING INDUCED EFFECTS WAVEFORMVS

4 DAMPED SINUSOID

I DECAYING EXPONENTIAL

218

I1AMPFfl SINlISO!I1L WAVFFOBN

v

WAVEFORM FREQUENCY tr Ins) id

I I Mz 12202 SO max Ampli1tude decreases',2 10 M Izt 201% 5 max 25-S06 in 4 cycles

ta L

+ + R2C- tdvo C R

1 MIz 10 Mmz

R :68 11 R45L=196 Jim 13.5 IuN

C - 129 pF C -18.7 PF

WAVEFORM t1. (ns) td (115)3500 mlax 170 (Nt

4100 max 2 (!2tV'i

j~k + +1 td

td 170 ps i., 2 t IO.

R =1.0 YD R 1.13 VuC s0.0068 iF 1 4/0 i

219

PRELIMINARY OBSERVATIONS

0 UPSET HAS BEEN INDUCED AND OBSERVED IN THE LABORATORY

8 UPSET DOES NOT OCCUR AFTER EACH TRANSIENT SIGNAL INJECTION

-4 IMPLIES CORRELATION BETWEEN UPSET AND PROCESSIN(P STATE

* NORMAL FUNCTION IS RESTORED BY RESETTING AND/OR

REPROGRAMMING THE SYSTEM

Q STATUS BIT SEQUENCES HAVE BEEN RECORDED THAT DO NOT

CORRESPOND TO AN 8080 MACHINE CYCLE

IMPLIES UNDEFINED PROCESSING' STATES BEING ENTERED

FUTURE PLANS

a COMPLETE THE FOLLOWING UPSET TEST MATRIX

INI I VUL- WAV1,IOIIM0 I DlATA DOS1 INES1

11 ClIIIP

WAVrT OflM 2 AUDlIESS PITS2 IINIS

x xWAVI:~10,A 3 CONTRlOL I.INIIS

CPU n Cl WAVEFOAiM 4 110 LINES

* PERFORM A STATISTICAL ANALYSIS OF THE TIME

DATA TO RELATE UPSET TO

0 INTERNAL STATE OF THE CAN .DIDATE PROCESSOR

I TRANSIENT SIGNAL INJECTION POINT

* UTILIZE A SOFTWARE SIMULATION APPROACH TO DEVELOP

FAULT SIGNATURES FROM THE STATUS BYTE DATA

220

DIAGNOSTIC EMULATION ANALYSIS--NEED AND TECHNIQUE

by

Mr. Gerard E. Migneault

Langley Research CenterNational Aeronautics and Space Administration

A brief description of the problem of determining failuremodes will clarify the usefulness of emulation. The dis-

cussion will cover the relationship of complexity, defini-tional flaws, specific software dependent behavior andlumped parameter analytical models. Deterministic and

stochastic application possibilities of emulation will beidentified, as will implementation details - at thelevel of a conceptual scheme and in terms of supportingcomponents. A sample application will be described.Future possible directions will be identified.

221

~"~ 7 ~r -

CHARACTERISTICS OF FAULT TOLERANCE

e REDUNDANCY

0 ERROR DETECTION

o CONTINUED OPERATION

-- RECOVERY

see INCREASED COMPLEXITY

* SAFETY REQUIREMENT IMPLIES PROBABILITY OF SYSTEM FAILURE

IN 10-HOUR FLIGHT LESS THAN ABOUT 10-9

e FAULT INTOLERANT SYSTEM

WORST COMPONENT/DEVICE

MTTF -1010 HOURS

NOT FEASIBLE TODAY

* SYSTEM MUST TOLERATE FAULTS AND BE MAINTAINED

222

N NA C I N- (N-1) A C x 142 R

i-Cl 1-1 0

SYSTEM FAILURE

NSYSTEM MTTF - a-DEVICE MTTF- j

a out aK C 7 Ck

FAULT TOLERANT SYSTEM

10 DEVICE SYSTEM MTTF+

MTTF

223

MAINTAINED FAULT TOLERANT SYSTEM

2A

*MTTF aL.~ 3 A lt

MAINT A J

-10 10 HOURS 1t

* UrT - DEVICE MTTF

EPAIR No. OF DEVICES

*ECONOMICS - MULIFUNCTION DEVICES

o MORE COMLEXITY

TOLERANCE OF DEFINITIONAL FLAWS SOFTWARE PRIME CULPRIT

N h

4 1-k 1-kt

(SYSTEM FAILIRE SSE ALR

TOLERANCE OF DEFINITIONAL FLAWS

SOFTWARE PRIME CULPRIT

224

NA C1 (N- I)A C 2 CN..R I

I- I- C C

-k I1tk 1k

SYSTM FAILURE

H/U +S/W FAULT TOLERANT MODEL

*ANALYTICAL. SOLUTION TECHNIQUE

PROS FAIL = 1-. [Au(I-k)]t N-R

*BUT

0.g9gggg(C 1 s10.9999<C 2 51

o!uL(1-k) 0.0000000001

MTTR ARECOVERY

0.1 SEC

.1 SEC

120M -

1200

120-

0 12

0.2

0 0.1 1 10 10 2 10~ 3 10 10 6

I kEkIN.ATION DIGITAL SIMULATION

TEST DURATION FOR10FAULT INSERTIONS FOR200 0 GATE EQUIVALENTS CPUWITH ARITRARY MEMORYANm SOFT WARE

225

OBJECTIVE

DEVELOPMENT AND SPECIFICATION OF AN EMULATION TECHNIQUE FOR GENERATINGSTATISTICALLY SIGNIFICANT QUANTITIES OF FAILURE MODES EFFECTS DATA OFHIGHLY RELIABLE COMPUTER SYSTEMS

JUSTIFICATION

e IN HIGHLY RELIABLE, FAULT TOLERANT SYSTEMS SYSTEM FAILURE MODES DUETO DESIGN FLAWS BECOME SIGNIFICANT - PERI4APS DOMINANT

e CREDIBILITY OF'ANALYTICAL RELIABILITY MODELS IS DEPENDENT UPON AMOUNTOF DATA FROM WHICH INPUT PARAMETERS ARE- DERIVED

• HIGH RELIABILITY (10"9) OF SYSTEMS BEING ANALYZED PRECLUDES USE/LIFETIMETESTING OF ACTUAL SYSTEMS BECAUSE OF INORDINATELY LONG KTTF's

* NO OTHER MEANS IDENTIFIED TO ACQUIRE SUFFICIENT DATA FOR THE ABOVE.

TECHNICAL APPROACH

e DEVELOPMENT OF FAST EMULATION ALGORITHMS CAPABLE OF SUPPORTINGFAULT INSERTION

-GATE LOGIC LEVEL- HYBRID LEVELS

GATE/CHIP/REGISTER/INSTRUCTION COMBINATIONS

e IMPLEMENTATION OF ALGORITHMS IN A HORIZONTALLY MICROPROGRAMMABLECOMPUTER AS AN ALGORITHM TESTRED WITH OPERATIONS SYSTEMS

* DEVELOPMENT AND SPECIFICATION OF NEEDED SUPPORT CAPABILITIES

- META COMPILERS/ASSEMBLERS- DATA RECORDING CAPABILITIES- RUN TIME CONTROL FUNCTIONS- FAULT TABLE GENERATORS- H/W DESCRIPTION TRANSLATOR- ENVIRONMENT SIMULATOR/INTERFACING- DATA POST PROCESSORS- OPERATOR GRAPHICS

226

-.. /WPLOGIC AcR~I~.-.

GENERATORO

INU HOGIC OBJECT CODE - -OUTPUTGENERATOR OI

IL DIAGNOSTIC O0BSERVER :: - ', DIAGNOSTICS.

A 0 0 0 0 1 11n 0 0 1 0 1 v

x 0 0 0 11 vv VV I 1 0 0 VY

:Az vz

New PriorState State

CONCEPTUAL EM1ULATIONI SCHEME

227

T.* '

V00

b e

d

c d fe

v6

a bc d etfa y3'

a I1100 00 0 000 0

0 00 1 100 0 00 00

d~~ 0 0 0 0 0 1 1 0 0 0 0

0 000110 0 0 0

( 0 0 00 00 0 0101

SAMPLE LATENT FAULT ANALYSIS

.3

FAILURERATrIO .2 \

0 1 2

''FCRETCCESBFR4ROEUSDT EEAE

'228

DIAGNOSTIC EMULATION APPLICATIONS

DETERMINISTIC:

I HARDWARE LOGIC DESIGN ANALYSIS.

I SOFTWARE DESIGN ANALYSIS,

I SYSTEM (H/W & S/W) EFFICIENCY/MISMATCH ANALYSIS#

I H/W/S/W/SYSTEM FAILURE MODES & EFFECTS ANALYSIS

I SYSTEM PERFORMANCE ANALYSIS IN A SIMULATED MISSION

DIAGNOSTIC EMULATION APPLICATIONS

STATISTICAL:

I LATENT FAULT ANALYSIS AND MODELING,

I COVERAGE DETERMINATION AND MODELING.

I TRANSIENT FAULT ANALYSIS.

229

DIGITAL SYSTEM HARDWARE DESCRIPTION TECHNIQUE USED IN EMULATION

by

Mr. Robert M. Thomas, Jr.

Langley Research CenterNational Aeronautics and Space Administration

A technique to translate a digital circuit descriptioninto the form required to emulate the circuit at the gate

and flip-flop level is described. This technique, imple-mented as computer programs, takes as input a descriptionof the integrated circuit and translates the descriptioninto tables for the emulation. Integrated circuit descrip-

tion language is described.

23123u APm ILAwI4W F1U

EMUSLATOR

--. 7ASSEMBLER T

FAULT

iI

GENERATOR

-~ INPUT HIW OBETCD - UPTGENERATOR LOGIC OBETCD- 1,UPT

EMULATOR

D ANOSTICOBSERVER IDAGOTC

OUTLINE

e HARDWARE DESCRIPTION LANGUAGE

o TRANSLATION

* FAULT GENERATION AND INSERTION

232

HARDWARE DESCRIPTION LANGUAGE

SYSI I MA I I C k 111 S ,I OR, I XIISI -tNti I I ARIWAIKI 106 1{

IN A FORM USABLE BY THE TRANSLATOR.

SOURCE OF INFORMATION

SCHEMATI CS

- INTEGRATED CIRCUIT TYPES- INTERCONNECTIONS- NAMES OF EACH IC

INTEGRATED CIRCUIT DATA SHEETS

- GATE/FLIP-FLOP MODELS

HARDWARE DESCRIPTION LANGUAGE

LANGUAGE ELEMENTS

GATE TYPES: AND, OR, NOT, NAND, NOR, EXCLUSIVE OR,EXCLUSIVE NOR

FLIP-FLOP TYPES: D, J-K, R-S, T

INTEGRATED CIRCUIT. IDENTIFIERS: ALPHA-NUMERIC

INTEGRATED CIRCUITCONNECTIONS: PIN TO PIN

233

CHIP DESCRIPTION

$ CHIP DEFINITIONTYPEPOWERDESCRIPTIONUNUSED PINSFUNCTIONS

G-S=G-AG-R=C-CG-A'

G-B'

G-D 0=G-D )G-D 2=G-D 3=G-D 4=G-D 5C-D 6C-D 7=G-W (P-6) =

G-Y (P-5) =END CHIP

LATCHED PARALLEL TO SERIAL CONVERTER SCHEfMATIC

DATA LATCH +Sv +Sv

0 DO DO 2 ~)

PARALLEL 2 D2 Q2 6 2 )2 y -- ~ SERIAL DATA3 - 03 Q3 9 1)3

DATA L 4 04 Q41; )I

014D

6 523 D A

4 12DS Q

7I CLOCKU CD7 010 IC)D

7 4 1 .5 2 7 3~ G H -- 1 0

SIROA~C>L - AlE MODELno 2 k.

D2A-

G- 04

D7~E L VCC = -1, G~D =

~~~~~~~HPDESCRIPTION FO 74EO IH AA EETR LLILS151

UNUSED PINlS N IONEFUNICTIN

-3= IOT(P-7)G-Al AIND P-i 1)G-B =Alit'(P-I 0)G-C a AND(P-9)6-A' = l4OT (G-A)'

II" OTtGB..-6-C' = IIOT(G-C)3-DO - AIID(6-S. P-4v 6-A-' * -B', 6-C.)G-DI - AllD(6-S, P-3v G-Aq G-B'o 6-C:')G-D2 - FIND (G-S, P-2, G-A's 6-B, G-C-)G-D3 - RflD(G-So P-I, 6-A, 6-B, 6-C')G-D4 - AND(.G-Sp P-15o G-A-'q G-B'o G-C)G-D-. - AHD (GLSZ, P-14w G-A, G-B', 6-C)G-D6 - AID (G-S 9 P-13o G-A>, G-B, 6-C)G-D7 - AtID'6-Sq P-129 6-A' G-Bp G-C)G-IWhP-6) a NOR(G-DO6-rI,6-D~,6-D3,6-D4,G6-D5,G-D6,G-D7)6-'v(P-5) a 110T6-IJ)

S EflD CHIP S

235

INTEGRATED CIRCUIT NAME TABLE

NAMEICJTYPE

IC 1 74LS273IC 2 74LS151IC 3 74LS169

INTEGRATED CIRCUIT CONNECTION TABLE

OUTPUT DESTINATION SIGNAL NAME(OPT IONAL)

IC 1-2 IC 2-4

IC 1-5 IC 2-3IC 1-6 IC 2-2Ic 1-9 IC 2-15

IC 1-12 IC 2-14IC 1-15 IC 2-13

IC 1-16 IC 2-12

IC 1-19i IC 2-5 OUTPUT CONNECTOR SERIAL DATA

IC31 C21

IC 3-13 IC 2-11IC 3-13 IC 2-10

236

*iiijl * 114

TRANSLATION PROCESS

TRANSLATION PROCESS

CONVERT A PSEUDO ENGLISH CIRCUIT DESCRIPTION TO APACKED BINARY ENCODED FORM REQUIRED BY THE EMULATIONALGORITHM

REJECT DESCRIPTION THAT DOES NOT FOLLOW PRECISELYTHE HARDWARE DESCRIPTION LANGUAGE RULES

237

EMULATION MATRIX

SOURCE

6-S 6-A Gil 6-C G-A' (i-f' (i-C' G-DO

6-A

6-B

6-C

6-A' x

DESTINATION 6-B' x

6-C'

6-DO x x x x

A 0 0 0 0 1 11 VA

x 0 0 0 1 1vy1I 1 0 0 -

Z V

z

New Prior*State State

CONCEPTUAL EM1ULATIONl SCHEME

238

FAULT GENERATION AND INSERTION

FAULT SELECTION

- MANUAL - SELECTED BY EXPERIMENTER

- AUTOMATIC - RANDOMLY SELECTED

FAULT INSERTION

- MODIFY EMULATION MATRIX TO REFLECTFAULT

FAULT OCCURRENCE

- INFORMATION TO EMULATION ALGORITHM

AS TO WHEN FAULT OCCURS

SUMMARY

DEVELOPING THE TOOLS TO SUPPORT GATE/FLIP-FLOP LEVEL

EMULATION OF DIGITAL COMPUTERS

i'1 - HARDWARE DESCRIPTIOIN LANGUAGE.i

- TRANSLATOR PROGRAM

- PAULT GENERATOR PROGRAMS

239

THE NEED FOR TRANSIENT DATA IN A

CARE III RELIABILITY ANALYSIS

by

Mr. Salvatore J. Bavuso

Langley Research Center

National Aeronautics and Space Administration

The recently developed CARE III (Computer-Aided Reli-ability Estimation) computer program incorporates atransient/intermittent fault model that is mathematicallyaccurate in contrast to many existing reliability evaluatorprograms which employ approximations for the transient/intermittent model. The CARE III transient/intermittent

model will be discussed and compared to existing approxi-mating models. Computational complexity arising from theuse of the CARE III transient/intermittent model will alsobe addressed.

'1

241 -

P5,a-a u

ADVANCED RELIABILITY ASSESSMENT TECHNIQUES

OBJECTIVE: DEVELOP A CAPABILITY TO ASSESS THE RELIABILITY OF ANY

FAULT-TOLERANT COMPUTER-BASED SYSTEM, INCLUDING

EXECUTIVE SOFTWARE

Interprocessor bus

Triplex computer architecture.

31='l X2 X3 - X1 X? X3 +X 1 X2 13(TVO Out or three channels operational)

so *X 1 X2 X3 2.At(01w out of three(A.Udwimls 'hahulo3 operatloalj)

*operational) )-p A2 tS 2 'l K1 2 3

-I3A(l ya

r3 :R X1

(syten failure)

Markov state space model of triplex channel RCS.

242

P0 (t + At).- P0 (t) - 3AC I tP0(t - 3u(1 - C1 )htp0 (t)

Urn t + At) - P0 (t), . dO~t -Up (t)At-O A

dPa -WOP(t)dt

-~~t . 3.C P0(t) - 12.PI(t)

dt~~ * ~ 2 1(t) - ).P2 (t)

dP3 (t) - 3X(1 - CI)P0 (t) +2,(1 - C 2 ) 1 t . 2 z

where P0 Is the probability ot the System being In state SO, that is

P 0 = P(50) P1 W P(s1 P2 a P(S2 ) P3 -

and the initial conditions are

P0 (0) . 1 - P .. P(O) P3(C) a0

P0 (t) e x

P1 (t) - 3C,(e-2xt -3)

P2t.3C C fe-A - 2e-2.t + 0-t)

PP() 1 - [P(t) + P1 1L) +. P(t)]

243

310 /0/0 m P/rWr2i

3

52

LAT 3ACC

-0-MI01.2S1.25

NORM 3ACC /.25/.25

3

3 0 /. 2S/.MPXA/D 2713

DG*

r3 S/.l COUPLERCOMPASS 2401.25/.40

190/.25/.40 Com. 293

IS/_oj

0- MZFROCm 5o/.25/.40v moiy l ..........................

U01.05/. 10 93

130 1. ZS/. 4'

3

31J-57-D&-j 3 DADS 3 PITOT 3 VRAM 20 200 STATIC

- - - - - 8013 300 /. 2S/. 40

ROLL-32130/.05/.l

WATCH 3 x 3$ 3PITC 3SERVO 1301.051.1

12/.05/.Io 3I HYDR. Sl 3 YAW 34

SUPPLY 60/.05/.10 SERVO 130/.OSI.l

CWS- DELETE R/A AND ILS

-Akw- Term ARCS Dependency Tree

244

6-1

MIN""

4--

0 I ih

49

o 0o

zz

4245

NUPIER OF STATE NEEDEDFOR'A:MAR0OVMODEL

LET: n-NUMBER OF *'COUPLED" STAGES

k1 NUMBER OF POSSIBLE FAULT TYPES IN M~ STAGE

NUMBER OF POSSIBLE FAULT STATES/TYPE 'IN f T1 STAGE

mi NUMBER OF MODULE FAILURES THAT CAN BE TOLERATED

IN THE iTH STAGE

THEN NUMBER OF STSTEN STATES N IS

n i L + j-I

R.I., If n -4 and k 2, A 3, u,2 for all it N 614,656

n

CARE III N 2fR (m + 1) 162

CARE III APPROACH

* DEFINE SYSTEM STATE ONLY IN TERMS OF

NUMBER OF EXISTING FAULTS

* INDEPENDENTLY EVALUATE TRANSITION PARAMETERS

AS A FUtICTION OF DISTRIBTION OF POSSIBLE

* FAULT TYPES AqD STATES

0DETERMINE RELIABILITY USING KOLM'OGOROV'S

FORWARD DIFFERENTIAL EQUATIONS

* NUMBER OF STATES DRASTICALLY REDUCED;

TRANSITION RATES NECESSARILY TIME-DEPENDENT 8 1 v)

246

9'7

COMPUTER AIDED RELIABIITY E S1I MATI OW

(CARE I II I

NII -ANYHIGIA (N"I' 11"

FAISTING~/C cr UA ON.

SYIICARE 1 1II&

c±J PROGRAM

F~AULT TREE

V ,FAU FAULT HANDLING

HANDLING DATA IPT

AOIPRSONA 8ONITOING

I.INTERMITTENSE A TANSoENTS 1 CYBER 175

F 4ut HANDCLASS

:ELECTRONIC COMPUTER*HYDRAULIC*MECH ANI CAL

110- WSAGES NST~tS-2.N1.5,M-3.3. IRLPCD-4,RLPLOT-TSle0- IU~CAT tWCAT8oi.lJTVPd1.1I)-l.JTVP(1,2)-1.

140- SRTI14E FT.1S. ,SYSFLO-TCPLFLG-TSISO- SIM1 FALT-RE 7-1-8I160- 1 2 3 3179- 3 0 1 a.I115- SIM1 CRITICAL-FAULT PAIRS190- 1 1S 16 18Me1 10210- 9 1S 1M.-162a12834 5 6?7S910

23-1? 2 11 12 13 14 1S240-18 0 16 1?

Sys.FAILUR

16 SYSTEM FAILS IF STAGEFAILU. 1 OR 2 FAILS OR A

30i 1 5BSFO FALROCCURS.

COMPUTER BusIsSAE STAGE 1A ''

1 2 (DSAGSYSTEM TREE 2/10 2/5 A CRITICAL PAIR FAULT MODEL

1-10 11-16

CRITICAL PAIR FAULT TREE

247

PERMANENT, TRANSIENT/INTERMITTENT FAULT ANALOG

PERMANENT FAULT TRANSIENT/INTERMITTENT FAULT

LOGICAL FAULT MODEL S-A-1/S-A-O S-A-I/S-A-OAFTER FAULT ARRIVAL DETERMINISTIC: T F < T <STOCHASTIC T FOR DURATION OF

S-A-i AND S-A-OEXPONENTIAL: f (0 = ae -i

FAULT ARRIVAL MODEL WEIBULL: f(t) = w)xtff e-Atw

EXPONENTIAL: f(t) = xe-At

TRANSIENT AND PERMANENT FAULTSCONVENTIONAL MARKOV MODEL

2-UNIT SYSTEM

A= PERMANENT FAILURE RATE (ARRIVAL RATE)

AT = TRANSIENT RATE (ARRIVAL RATE)

1= PROPORTION OF TRANSIENTS THAT ARE MISTAKEN FOR PERM4ANENT FAILURES

I- (2Xp+ 1- 21TIA 1 x + 1 .xT )At

(2AT1)T)t(1 t.i

01F

248

C9 F

B E

Overall reliability model of two-unit system.

Aggregated reliability model for a two-unit system.

System states defined by number of failed nodulesof each type

* Transition rates determined by averaging overallfailure types and states

249

SINGLE-FAULT MODEL EQUATIONSVARIABLE

*(t) 3 - 0 trUOd(r FEEAR( IT)

ftP t -. rt~)+ 0 * (t-T) P (-r)dT- PAI(IT)

Pb (t) ( *t. + *k)brd 1 (T)

p~t e M *ap(T)d()e(t-T)dT + * (t-r)P eCT) d- PERR(IT)

PS (t) - e 0'(t)d(t) + 0 3 (t-t)p e(T) dr PEARCIT)

p.(t) - t Wr (t) + 0 ) *t-Tr)p-(r)dT PNEAR( IT)

,t

Pf(t) -(1-C) pe Ur) E t-r) dT PFLD( IT)

TA () 0 C Pe(T)C(t-T)( + ae 4 8 )C dT + pwCt) PSIA(IT)

BW 1T- Pe()(

250

oolib) Fau] t Moud!

02 p 1 (t)

A 11

IA 6 (t)

0 OD2

A

2

ADVANCED RELIABILITY ASSESSMENT DEVELOPMENT

ACCOMPLISHED 1981 1982 1983

DEVELOPED RELIABILITY CARE IIIASSESSMENT PROGRAM - A ENHANCEMENTBASIC TOOL FOR THE AUTO SENSITIV;COMPARATIVE ANALYSIS OF QUICK LOOK

DIFFERENT AVIONIC SYSTEMS. CARE IIIVALI DAT ION

APPLICABILITYo ARCHITECTURAL SOFTWARE

STRUCTURE4 FAULT TOLERANT- NUIIR Cl

M ECHANISMS CARE I I Io TRANSI ENT ANALYSIS

TOLERANT TECHNIQUEMECHANISMS

o FAILURE RATE W IERCEliDATA PUR HRETICA

a AVIONIC ICA

SUCCESS IDATIONCRITERIA

PEER REVIEW

251