Shipboard wave height sensor, Atwater, R. 1990

(1)

SSC-362

SHIPBOARD WAVE HEIGHT

SENSOR

This &cumenl has been approved for public release and sale; its

distribution is unlimited

SHIP STRUCTURE COMMITTEE

(2)

The SHIP STRUCTURE COMMITTEE is constituted to prosecute a research program to improve the hull structures of ships and other marine structures by an extension of knowledge pertaining to design, materials, and methods of construction.

RADM J. D. Sipes, USCG, (Chairman) Chief, Office of Marine Safety, Security

and Environmental Protection U. S. Coast Guard

Mr. Alexander Malakhoff Director, Structural Integrity

Subgroup (SEA 55Y) Naval Sea Systems Command Dr. Donald Liu

Senior Vice President American Sureau of Shipping

Mr. Albert J. Attermeyer Mr. Michael W. Tourna Mr. Jeffery E. Beach MARITIME ADMINISTRATION Mr. Frederick Seibold Mr. Norman O. Hammer Mr. Chao H. Lin Dr. Walter M. Maclean

SHIP STRUCTURE COMMITTEE

Mr. H. T. Haller

Associate Administrator for Ship-building and Ship Operations Maritime Administration Mr. Thomas W. Allen Engineering Officer (N7) Military Sealift Command

CDR Michael K. Parmelee, USCG, Secretary, Ship Structure Committee U. S. Coast Guard

CONTRACTING OFFICER TECHNICAL REPRESENTATIVES

The SHIP STRUCTURE SUBCOMMITTEE acts for the Ship Structure Committee on technical matters by providing technical coordination for determinating the goals and objectives of the program and by evaluating and interpreting the results in terms of structural design, construction, and operation.

AMERICAN BUREAU OF SHIPPING Mr. Stephen G. Arntson (Chairman) Mr. John F. Conlon

Mr. William Hanzalek Mr. Philip G. Rynn

MILITARY SEALIFT COMMAND

U. S. COAST GUARD ACADEMY LT Bruce Mustain

U. S. MERCHANT MARINE ACADEMY Dr. C. B. Kirn

U.S. NAVAL ACADEMY Dr. Ramswar Bhattacharyya

STATE UNIVERSITY OF NEW YORK

MARITIME COLLEE

Dr. W. R. Porter

WELDING RESEARCH COUNCIL Dr. Martin Prager

NAVAL SEA SYSTEMS COMMAND Mr. Robert A. Sielski Mr. Charles L. Null Mr. W. Thomas Packard Mr. Allen H. Engle U. S. COAST GUARD CAPT T. E. Thompson CAPT Donald 5. Jensen CDR Mark E. NoII

SHIP STRUCTURE SUBCOMMITTEE LIAISON MEMBERS

NATIONAL ACADEMY OF SCIENCES -MARINE BOARD

Mr. Alexander S. Stavovy

NATIONAL ACADEMY OF SCIENCES -COMMITTEE ON MARINE STRUCTURES Mr. Stanley G. Stiansen

SOÇJETYOENA\LABÇtIIIEÇ[$..AND

MARINE ENGINEERS

-HYDRODYNAMICS COMMITTEE Dr. William Sandberg

AMERICAN IRON AND STEELJNSTITUT Mr. Alexander D. Wilson

Mr. William J. Siekierka Mr. Greg D. Woods

SEA55Y3 SEA 55Y3

Naval Sea Systems Command Naval Sea Systems Command

(3)

Mcmber Agencies:

United Statss Coast Guard Naval Sea Systems Command Maritime Administration American Bureau of Shippírig MilitarySealiftCommand

Ship

Structure

Committee

An Interagency Advisory Committee Dedicated to the Improvement of Marine Structures

December 3, 1990

SHIPBOARD WAVE HEIGHT SENSOR

The ability to obtain accurate ship motion information and to

correlate these data with local sea conditions are necessary for advanced ship and ship motion research. This report describes

the development, testing, and assessment of a prototype system designed to simultaneously acquire ship motion and wave height

data. The integrated system includes a pulsed laser wave height

sensor, vertical accelerometers, motion sensors for roll and

pitch, and a data acquisition computer. The system was installed

on an ocean going vessel for testing and evaluation. The reliability and accuracy of the prototype system were considered

as well as the robustness or survivability of the unit under adverse sea conditions.

.3 D SIPE

Rear Admiral, U.S. Coast Guard Chairman, Ship Structure Committee

c3C

- 2..

Address Correspondence to:

Secretary, Ship Structure Commthee U.S. Coast Guard (G-Mm)

2100 Second Street SW. Washington, D.C. 20593-0001 PH: (202) 267-0003 FAX: (202) 267-0025 SSC-362 SR- 13 14

(4)

Technical Report Documentation Page

L Report No.

SSC-362

2. Government Accession No. 3. Recipient s Catolog No.

4 Title and Scbtitle

SHIPBOARD WAVE HEIGHT SENSOR

5. Report Dote

November 1990

6. _{Perforr,rrng Organization Code}

8. Performing Organization Report No

SR 4314

7. Aothor1s)

R. Atwater

9. _{Perforr,rrng Organization Nome ond Address}

Scientific Applications International Corporation

1522 Cook Place Goleta, CA 93117

10. _{Work Unit No. (TRAIS)} 11. _{Contract or Grant No.}

DTCG23-87--2 O

13. Type of Report and PeriodCovered

Final Report

12. Sponsoring Agency Narrte and Address

Ship Structure Committee

U. S. Coast Guard

2100 Second Street, SW

Washington, DC 20593

i

14, _{Sponsoring Agency Code}

GM

15. Supplementary Notes

This work was sponsored by the Ship Structure _{Committee and its member agencies.}

16. Abstruct

This report summarizes the results of an effort to develop, test and evaluate a Pulsed Laser Based, _{Shipboard Mounted, Ship Motion}

Compensated Wave Height Sensor. _{This system was used in a North} Atlantic run to measure sea surface waves at normal sea speed. The

wave height sensor utilizes an infrared wave surface range sensor coupled with a vertical accelerometer and pitch and roll sensors. The

prototype system was evaluated for measurement accuracy and

reliability as well as ruggedness and survivability in elevated sea

state conditions. _{Lack of ground truth precluded a comprehensive} qualitative evaluation of overall performance. _{Data drop outs were} experienced when operating in elevated seas thereby making further work necessary on the sensor and on the data processing software. Prototype system characteristics, _{signal processing flow chart and} analyzed data are presented.

As an adjunct to the development of the wave height sensor, the Ship Structure Committee and the Society of Naval Architects and Marine Engineers cosponsored a project at the University of Michigan _to investigate through model testing the influence of ship generated waves in front of the bow. _{The report of this study is}

included as Appendix D.

17.

ii'otion Sea State Measure Ship Response _{Ship Accelerations} Wave Height Roll

Pulsed Laser Sensor Pitch

Bow Displacement Heave

18. Drstrrbutron Statement

Available to the public from: _{National Technical Information} Service, Springfield, VA 22161 or Marine Technical Information Facility, National Maritime Research Center,

Kings Point, NY 10024-1699

19. Security Classif. (of this report) UNCLASSIFIED

20. _{Security Classif. (of this page)}

UNCLASSIFIED

21. No. of Pages

98

22. Price

Form DOT F 1700.7 (8-72)

(5)

LENGTH

(2000 1W

VOLUME

METRIC CONVERSION FACTORS

9

-23

Approximate Conversions to M.tr(c Measures

Approximate Conversions from Metric Measures

22 Symbol When You Know Multiply by To Find Symbol 2 Symbol When You Know Multiply by To Find Symbol 8 -

-LENGTH AREA cm2 square centimeters in2 square meters km2 square kilometers ha h.ctatei (10.000 m2) MASS (weig.fl) tsp tSisÇOOfl$ 6 mIlliliters ml Tbsp tablespoons 15 millilIters ml fI oz fluid OUACI 30 milliliters ml C P5 0.24 liters I pt pInts 0.47 liten t qt quarts 0.96 liters I gal galIons 3.8 liters I h) cubic feet 0.03 cubic meter, '2 yd3 cubIc yards 0.76 cvb4c meters in1 TEMPERATURE (exact) s FahrenheIt 5/9 (after Ce3.js O( temperature subtracting temperature 32)

'lin. - 2.54cm (erectly). For othat erect conversione end mor, detall tablee NIe Misc. Pubi. 218. UnIt, of W.lit and Meseures. PrIce $2. 25 90

Ctalog No. C13 10216. ml milliliters 0.03 fluid ounces fi o. I liters 2.1 nts I liters 1.06 quant ql I liters 0.28 gaRons gel m1 cubic meters 36 cubic feet ft3 m3 cubic miter, 1.3 c*jbic yards TEMPERATURE (exact) O Celsiut 9/5 (thin Fahrenheit temperature add 32) temperatura °F °F 32 96.6 212 -40 0 140 9f) I 120 160 2001 i I ,J l I I 4 I I I Ç I I ¿ t. I t I -40 -20 20 40 60 80 lOO °C 37 Itt Inch.. 2.5 csntlmetsrs cm ft let 30 c.citsm.t.rI cm yd yerds 0.9 m ivi milis 1.6 kilometers km AREA in1 squat. inch.s 6.6 quirs cantimetati cm2 ha yd2

ei.. f

i.

squats yards 0.00 0.8 meter, squat. meurs in2 m squat, mli.. 2.6 square kilomstsra km2 sae. 0.4 h.clar.s ha MASS (weit) O. ounc.. 29 smna 9 b pounds 0.46 kiIome kg sfiori tons 0.9 tonni. t mill,m.tirs 0_04 lnch.s in c.ntim.t.r. 0.4 Inchet In meurs 3.3 feet ft miter, 1.1 yards yd kilomteqs 0.6 miii. mi 0.16 square inc*iet In2 1.2 squara yards yd2 0.4 square miles ml1 2.5 acre. ---.19

-

_-= E 16

-

= li 18 6 -E 14 -E 13 5 12 11 4 = =

-10 9 E 8 3

:

--

E

:

6 -2 5 4 E I -E -2

-i E ¿tiches cm g erns 0.036 ounces o' kg kilograms 2.2 pounds b t tonnes (1000 kg) 1.1 sfiori toni VOLUME mm n m m km

(6)

TABLE OF CONTENTS

SECT!ON TITLE PAGE

1.0 SUMMARY i 1.1 Measurement Reliability/Accuracy i 1.2 System Ruggedness/Survivability 4 2.0 BACKGROUND 5 3.0 PROGRAM OVERVIEW 5 3.1 Phase I 6 3.2 Phase II 7 4.0 PROTOTYPE SYSTEM 9

4.1 Bow Mounted Sensors 11

4.2 Gyro Pitch/Roll Sensors 11

4.3 A/D Conversion System 11

4.4 Computer System and Software 14

5.0 PERFORMANCE EVALUATION 14

5 1 Angle Measurement Subsystem Performance 16

5.2 Vertical Acceleration Sensor Subsystem Performance 24

5.3 Wave Surface Range Sensor Performance Evaluation 27

5.3.1 Data Drop Out Detection Algorithm 29

5.3.2 Wave Surface Range Reconstruction Algorithm 57

5.4 Overall Performance 59

6.0 RECOMMENDATIONS 61

REFERENCES APPENDICES

LIST OF TABLES

TABLE # TITLE PAGE

i Gyro/Pendulum Pitch Roll Sensor Comparison 17

2 Drop Out Detection Statistics 36

3 Comparison with NOAA Buoy #41002 Data 60

(7)

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

ii

Parameter Parameter Parameter Parameter LIST OF FIGURES

FIGURE # TITLE PAGE

USNS DENEBOLA

Wave Sensor Installed on USNS DENEBOLA

Block Diagram - Prototype Shipboard Waveheight System EMI Wave Gauge Sensor Assembly

SSC Sea Trial Vessel - USNS DENEBOLA (SL-7) Overall Flowchart for Signal Processing

Gyro/Pendulum Comparison File: 131800.DAT Gyro/Pendulum Comparison File: 16220.DAT Gyro/Pendulum Comparison File: 083801 .DAT Gyro/Pendulum Comparison File: @1 6HzO. DAT Gyro/Pendulum Comparison File: 17550.DAT

Parameter Data Using Pendulum Sensors File: 080301.DAT

Parameter Data Using Pendulum Sensors File: 083801.DAT

Flat Spot Detection/Reconstruction

Pitch and EMI Range File: 08380l.DAT Pitch and EMI Range File: @l6HzO.DAT

Pitch and EMI Range File: l2590.DAT Pitch and EMI Range File: 134200.DAT Drop Out Severity vs. Pitch

Sea Trial Data File: 131800.DAT

Sea Trial Data File: l6220.DAT

Sea Trial Data File: l34200.DAT

Sea Trial Data File: 083801 .DAT

Sea Trial Data File: 08220.DAT Sea Trial Data File: @l6HzO.DAT Sea Trial Data File: 08380.DAT

2 8 10 11 13 15 18 19 20 21 22 25 26 30 31 32 33 34 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 . 54 55 56

Time Series (Non-Flat Spot Reconstructed) File: 131800.DAT Time Series (Non-Flat Spot Reconstructed) File: 08380 1.DAT Time Series (Non-Flat Spot Reconstructed) File: @l6HzO. DAT Time Series (Non-Flat Spot Reconstructed) File: 17550. DAT

(8)

1.0 SUMMARY

This report summarizes the development and evaluation of a pulsed laser wave height sensor

suitable for use on an underway vessel operating in an elevated seaway.

A prototype

measurement system based on an infrared wave surface range sensor, coupled with a vertical accelerometer and pitch and roll sensors was evaluated while the system was mounted on the bow of the USNS DENEBOLA (SL-7) en route from Bremerhaven, Germany to Savannah, Georgia during October 1988. Figure 1 shows the sensor and DENEBOLA operating in sea state 4 on October 23rd. Data obtained during this sea trial are discussed and provide the foundation for recommended future work.

The prototype system was evaluated from two points of view: (1) accuracy and reliability of

wave height measurements; and (2) ruggedness/survivability under elevated sea state conditions. The issue of wave field contamination from vessel-generated waves was considered analytically during the initial design phase.

1.1 Measurement Reliability/Accuracy

System performance was evaluated in terms of three major subsystems: (1) the gyro and

pendulum pitch and roll sensors; (2) the vertical acceleration sensor and vertical bow

displacement calculations; and (3) the infrared wave surface range sensor. The pendulum pitch and roll sensor data were found to contain a significant response to the vertical acceleration of the vessel. For larger accelerations, as much as a degree or two was added to the measured pendulum pitch and roll angles when compared with the more accurate gyro sensor data. The

effects of the errors in angle measurements, while not significant for the encountered sea states,

will be more severe for higher seas (larger vertical accelerations) and may compromise the

calculated wave heights when pitch and roll angles are more extreme.

(9)

Figure 1

USNS DENEBOLA

1346 October 23, 1988

7.3 feet Significant Wave Height

(10)

The vertical accelerometer and bow displacement calculation algorithm performed well; no significant errors related to data sampling, filtering or integration were evident in the computed bow displacement data.

The infrared laser wave sensor exhibited a response problem under certain sea state conditions. As discussed in Section 5.3, the signal from this sensor would intermittently "hold" at a constant value for intervals of a few seconds during conditions of high pitch motion. The onset of these

"data drop outs" occurred shortly after the minimums and maximums of the pitch signal.

Examination of the data indicates that the severity of the drop out problem is dependent on both the amplitude and period of the pitch motion. A data reconstruction algorithm was developed that appears to adequately interpolate through the data drop out intervals. By eliminating the drop outs, the reconstructed time series, associated spectra, and computed statistics are consistent

with visual observations, model data and buoy measurements.

The "Recommendations" section of this report provides suggestions for further evaluation and

possibly limiting the effects of the data drop outs. Such possibilities: include (1) sensor

malfunction under the more extreme sea-states; (2) acquiring "raw" sensor data instead of data which has been screened by the logic circuitry of the wave gauge; and (3) further investigating

whether the cause of the drop outs is "bow splash" by obtaining data while the sensor is

temporarily mounted away from the bow.

The lack of ground truth information for the prototype sea trials precludes a comprehensive quantitative evaluation of overall performance of the wave height measurement system. In this regard, the results of the sea trials must be considered somewhat inconclusive, particularly with respect to the ability of the infrared laser to operate in the presence of vessel induced waves or spray. The data drop outs limit the reliable performance envelope of the prototype. However,

revisions and improvements to the electronic signal rejection logic and data reconstruction

algorithms may overcome these limitations.

(11)

An improved understanding of the drop out phenomena combined with a more rigorous ground truth experiment will provide a definitive evaluation of the operational utility of the pulsed laser wave height system for underway measurements.

1.2 System Ruggçdness/Survivability

The prototype system consists of three physically separated groups of components:

The wave gauge, vertical accelerometer and pendulum sensors were packaged in an explosion proof housing mounted at the bow of the DENEBOLA, No damage

occurred to any of these sensors or their mountings during either leg of the

voyage. The system was installed but not operating during the eastward crossing,

and was operating continuously during the westward crossing. Elevated sea

states, including "green water" over the bow, were encountered during the

eastward crossing. Sea states were milder during the westward crossing.

The gyro-type angle sensors were installed in the wheel house. No problems were experienced by these sensors throughout the sea trials.

The data acquisition computer system performed without any major problems. The system operator's notes report occasional questionable computer performance, and a jammed printer ribbon, but these are deemed minor "inconveniences" rather

than ruggedness/survivability problems. The uninterruptable power supply (UPS) for the computer was not functional during the sea trials. The UPS was stored in the wheelhouse near an open door during the eastward leg of the voyage and may have been damaged by exposure to salt air during this time.

(12)

2.0 BACKGROUND

An accurate analysis of full scale ship performance is dependent upon a reliable measure of the encountered seaway. Researchers have repeatedly been frustrated by the lack of a reliable and accurate sensor suitable for underway measurements. As part of the SL-7 and the STEWART J. CURT research programs, the Ship Structure Committee (SSC) and the US Coast Guard have

evaluated a variety of shipboard wave height sensors. Based upon the results of these

evaluations, Dalzell (Ref 1, 2) concluded that neither the Tucker pressure meter nor the OWHS

radar system were suitable

for reliable measurements in an elevated seaway. The

recommendations presented in SSC Report #313 (Ref 3) indicated that the development of an improved shipboard wave height measurement system was critical to future full scale vessel research. Subsequently, the NMRC (Ref 4) analyzed the system requirements and application of a pulsed laser for underway measurements. The results of the NMRC work provided the impetus and direction for the current shipboard wave height sensor development program.

A variety of problems have been encountered with previous shipboard wave measurement

systems. The Tucker wave meter has been used relatively successfully on stationary weather ships; however, because of non-linearities related to vessel speed, it is not suitable for underway measurements. Narrow beam radar altimeters have suffered structural damage when mounted on exposed bow locations. Moving the sensor to more protected locations compounds the problem of removing the ship motion effects. Also, radar, microwave and sonic sensors are relatively wide aperture sensors that loose definition of the individual encountered wave shape. Previous investigators also reported considerable difficulty eliminating noise and drift errors from the double integration of accelerometer data used to compensate for vessel motions.

3.0 PROGRAM OVERVIEW

The objective of this program was to develop, test and evaluate a pulsed laser wave height

system suitable for installation on the bow of an ocean-going vessel. The program consisted of two phases as shown below:

(13)

Page 6

PHASE I

o Analyze Sources of Error

o Simulate System Performance

o Breadboard System

o Breadboard Sea Trial

o Develop Prototype Design

PHASE II

o Prototypes System Development

o Ocean Test

o Data Analysis

3.1 Phase I

The Phase I objectives were satisfied through a combination of analytical studies, computer simulations and breadboard experiment. The primary issues considered during Phase I were the effect of wave contamination, laser inclination, sensor stabilization and accelerometer integration errors. The results of these investigations were used to configure a breadboard system

demonstration and refine the prototype system design. A complete discussion of the Phase E, work is presented in the SAIC Phase I report (Ref 5). The significant conclusions derived from Phase I are sunirnarized below.

o Wave contamination is a potential problem for any shipboard sensor. At elevated sea states, vessel generated waves may reduce the measurement accuracy.

o Based upon breadboard experiments, the pulsed laser sensor should operate

(14)

o For the vessel motions typical of the SL-7 in sea states

..

7 a gimbaled sensor platform is not required.

o The acceleration should be sampled at 16 Hz and digitally integrated to

minimize integration errors and preserve phase coherence.

3.2 Phase II

The prototype wave measurement system design developed during Phase I was built and tested

during Phase II. The system includes a pulsed laser wave height sensor combined with

appropriate vessel motion sensors and a real time computer data acquisition and display system. An EMI pulsed laser wave height sensor mounted in a ruggedized explosion proof housing was selected for the prototype. This sensor has been successfully used on fixed offshore platforms since 1982. Selecting an existing and proven sensor sub-system minimized the development

effort required to validate the application of infrared laser technology to the shipboard wave

measurement requirement.

Upon completion of laboratory testing, the prototype system was installed on the bow of the USNS DENEBOLA (Figure 2) for test and evaluation during the October 1988 Atlantic

Crossings. The ruggedness of the wave sensor and installation were verified on the east bound

crossing when "green water" was taken over the bow. Data acquisition was successfully

conducted during the west bound trip.

Approximately 60 wave measurement data sets were acquired during the voyage from

Bremerhaven, Germany to Savannah, Georgia. Each data set included measurements of wave height along with vessel motions. Visual observations of wind and wave conditions were also recorded.

Measured ground truth data was limited to a NOAA weather buoy located

approximately 250 miles east of Charleston, SC.

(15)

Figure 2

(16)

Data obtained during the October sea-trials from the underway wave measurement system has been analyzed relative to sub-system performance and overall wave measurement capability. Because of the limited ground truth, and relatively mild conditions, the sea trial results are not totally conclusive. As anticipated, wave spray adversely effect the laser beam. Data drop outs from spray or excessive sensor inclinations are partially overcome with an adaptive data recovery algorithm. The processed data witb ør without occasional drop outs compares favorably with

the observed wave conditions.

With additional work to minimize data dropouts and/or adaptively interpret between good data, the pulsed infrared laser will provide a robust sensor for underway wave measurements in an elevated seaway. The following sections of this report discuss the prototype wave measuring system and the results of the October 1988 sea trials.

4.0 PROTOTYPE SYSTEM

The initial proof-of-concept "breadboard" system design was tested in the Santa Barbara channel as described in the SAIC "Phase I Report, Shipboard Wave Height Sensor" (Ref 5). The Phase I Report summarized the objectives and environment considerations for the system, and described the breadboard system used to test the sensor and processing components of the to-be-built prototype system. The results of the breadboard system trial were judged encouraging, within the limitations of the low sea state in effect when the test was conducted.

The prototype system block diagram is presented in Figure 3. This system can be divided into four major subsystems: (I) a sensor subsystem containing the EMI infrared laser wave surface range sensor, vertical accelerometer, and pendulum-type pitch and roll sensors; (2) separately positioned gyro-type pitch and roll sensors; (3) an analog-to-digital conversion subsystem; and (4) a PC/AT compatible computer system with attached monochrome display and hardcopy printer. These subsystems are described in the following sections.

(17)

EXPLOSION PROOF WAVEHEIGHT

HOUSING

EMI #2 00/1630/S

mfraRed Waveheight

Sensor

Schaevitz #2383-01 Pitch Clinometer Schaevitz #2383-01

Roll Clinometer

Schaevitz #LSBP-2g

Vertical

Accelerometer J.E.T. #VG2O4F

Vertical Displacement

Gyro (Pitch/Roll)

PowerOne HP15-1.5a +15 V Power Supply

JET. #PSC-100A Converter! D e ni od

Metra Byte #31 02

Signal

Conditioning M od u les

PowerOne HTAA-16WA +15 V, +5 V Power Supply

Sola #750 VA

UPS

Everex #1700

PC/AT Computer w/Co - Processor

Metra.:Byte OS-8

MUX A/D

Converter

40 M Byte Hard Disk 12'M Byte _{Floppy Disk} 60 M Byte

Cartridge Drive

Figure 3

Block Diagram - Prototype Shipboard Waveheight System

DISPLAY

Okidata #182

Printer

(18)

-Page 11

r

_{4. 1} _{Bow Mounted Sensors}

Figure 4 shows the EMI wave gauge sensor assembly. The pitch/roll vertical accelerometer and pendulum sensors were mounted inside the wave sensor an explosion proof steel housing. The explosion proof housing was bolted to a steel mounting tray and the mounting tray was bolted to a steel housing "box". At installation, the housing box was welded to the bow plate of the USNS DENEBOLA. The downward look angle of the wave sensor was variable by adjustment

of the bolts fastening the mounting tray to the housing box. The initial angle was set at 12

degrees and was not changed during the sea trials. The USNS DENEBOLA geometry

installation is shown in Figure 5.

A waterproof cable provided power to the EMI wave sensor and vertical accelerometer from DC power supplies in the wheelhouse , and returned the sensor signals to the AID conversion system

mounted on the backplane of the computer.

4.2 Gyro Pitch/Roll Sensors

The gyro pitch/roll sensors were mounted in the wheelhouse in a vibration damped housing; signals from these sensors were cabled to the A/D converter.

4.3 A/D Conversion System

The analog sensor signals were converted to digital form by a 12 bit A/D converter mountedon the backplane of the computer. Signal conversions were rapidly clocked and multiplexed by the A/D subsystem so that inter-channel timing skew was minimized. The interval between A/D scans was set to an operator-determined sample rate controlled by a software timing loop in the data acquisition program. Scan intervals of 8 and 16 Hz were used during the sea trials.

(19)

1ALLAJ19.-) W6

t) »n.

.1 04 sr

S)

7-a Ces a-'.

*.. 11') 4 se-0 :fl. 2t -4 Q'.i.a ana tu.4 or 404 '-60 Ca b-.'.- ta 7-55219 0) .onCO.45.'.a).. 4$60-.a.en40n)e ._ *4a14 .t4 -. * *-. 5*0 * St't 4* * *_. a...n,a.oIrn.o-o.. 4) ....C.J (rtM S) 4 60.6.7-ô fl* PG) 44 045 o' I'S 1 _'.(0 S*0* -4 r42TAPI 4TE' an-os * 5*t,*cO-'--'S C..1O, nn_ ,Ss,.*Ì,ba60 --40? St

c..,ca 'O 00 T4?* IH-,**t-60ThST,-a bosar& c*fl00-1%

t4 1* t tna6-) A. F.-n_ 04540*5 A-..-*0o4 4)021%. 5) *OOS t" S) T'. 64 ':440401*4 1 O'15-':, 4 01*4 (440* .1) 00 00405 S CS*0 1-a 00440.46 t ais-si-24 '-4fb ('SS-a 06« 7-4 1* 1 fOr 1.6 ss.. 1.qs:4400_46 Ct-stb'. Q54 414 4 ttt t 540*0 40*4. '4* 5, 0* .10.4.44 7*0 *1'* 44., .4. .S0O 44*. ('S 7-440.250_6 ':.'q 1-s 4 14*1 0 40*0*0*..':.. 5:44:4 lOOT 7-fl'. S-tb 500! .60..r,44, S..-(.'b*s .4) 4..o*..tl" Is Ssst..-t... -s. M' 4) '.' .... -1.6 p.... r'... .._*o I..' l' ..*..*.l r.'.. TOo) (T7-°> S) '::l 0.fl as--25 so.... *0-005 34 S-405104 05*0250 4 .&'4 O 4*C4as.164... 4025 aity- 24-.,. .0a(Go *044S-'. T-. s .444*404 40*4405.. MS) 0*0(.-oj S).?'. S),Io 4. Ib

4q.-

Sto,.'..' -4

o o e

ej j 44.4 12416 IT o.. 6) LaAlt ,...,ç. It.. 00.. .404.4! 0W oPP.'O S*P LS t-4.24 t .1*4'. '20)-4540204 L_4.24. 00.

Figure 4

10*0

»'

Oc 1644.1 I(.'-'T 561250Q

N

0441 0*4*07.04 P204flâ. i 2G0/1630

(20)

VCG 40' ABV B/L

LENGTH OVERALL 946' _o 10 20 30 40 50

SCALE: FEET

Figure 5

SSC Sea Trial Vessel

USNS DENEBOLA (SL-7)

55'

(21)

4.4 Computer System and Software

The real time sensor data acquired by the A/D conversion system was processed by a computer program running on the Everex 1700 PC/AT compatible computer system. This computer has

dual clock frequency capability, switchable to 8 or 12 MHz. The 12 MHz setting was used

during the sea trials.

The overall logic of the real time processing program is shown in Figure 6. As shown by this figure, the processing sequence consisted of several steps: (1) A/D conversion and scaling to physical units; (2) correction of measured acceleration and wave surface range to the vertical;

(3) double integration of the corrected acceleration to calculate bow displacement; (4) decimation from the 8 or 16 Hz sample frequency to the 2 Hz processing frequency; (5) wave height com-putation (wave range less bow elevation); (6) statistics comcom-putation (mm, max, mean, RMS, zero crossing period and max peak-to-peak cycle); and (6) data display (raw time series, processed time series, spectra, and statistics time histories). Appendix A lists the processing modules which perform the functions described above and details the double integration/filtering scheme ('TRANSFORM' flowchart) used to compute bow displacement from vertical acceleration.

In addition to the real time processing and display sequence listed above, at the operator's

discretion both raw and/or processed data could be stored on the hard disc for subsequent re-processing. A tabulation of data recorded during the October 1988 sea trials is found in

Appendix B; this appendix also describes the format of the data records.

5.0 PERFORMANCE EVALUATION

This section of the report discusses the performance of the prototype wave height measurement

system in terms of the three major subsystems: the angular position subsystems; the heave

measurement subsystem; and the wave surface range measurement subsystem. The overall performance of the system in terms of the accuracy and reliability of the under way wave height

measurements is discussed in paragraph 5.4.

(22)

C ON V RT TRENDS EMI EMI Range A BACK TO TOP FOR NEXT CYCLE

STORED AWAY AND AVERAGED WITH PSD DATAFROM PREVIOUS CYCLES FOR DISPLAY.

CON V RT TRENDS ¿ Accel TRIG L EMI CON V FU TRENDS Roll Bow Accel TRANSFORM y Pitch Bow Velocity (stored) TRANSFORM Bow Displacement Wave Height y Dis pl. See Separate Flowchart Bow Dis pl. y y DECIMATE Bow Veloc. Bow Bow Veloc. Accel. Bow Accel. V

STATS AND STATS DISPLAY

Figure 6

Overall Flowchart for Signal Processing

Roll Roll y y Pitch Pitch PSD Page 15 EMIDISPL. Wave Height

Wave Bow Bow _Bow _Roll _Pitch

Height Displ. Veloc. Accel.

¿ ¿ ¿ y y V

y ir y y

PITCH

GYRO _PENDULUMPITCH

ROLL PENDULUM ROLL GYRO BOW ACCELERATION EMI RANGE

(23)

The data discussed in this section were obtained over the time period from 19-25 October, 1988

the USNS DENEBOLA (SL-7), was en route from Bremerhaven, Germany to Savannah,

Georgia, crossing the Atlantic at a heading of about 230-265 degrees at speeds ranging from

under 10 knots to over 20 knots. Over 60 data files containing the raw sensor data were

recorded during the sea trials. Each file contains the unprocessed output of the AID converter for the pendulum and gyro pitch and roll sensors, the vertical accelerometer, and the EMI wave

gauge. These files are listed in Appendix B. A representative sample of the sea trial data, covering the range of pitch, roll, acceleration, and wave height measurements was reviewed while preparing this section of the report. Selected records were re-processed from the recorded

raw data using a modified version of the real time software.

As discussed below, the

modifications included: (1) a geometric correction for the tilt of the laser/accelerometer sensor; and (2) an algorithm to detect "data drop-outs" in the wave gauge signal and to estimate the true

wave surface range during periods of missing range data.

5. 1 Angle Measurement Subsystem Performance

Pitch and roll angles were measured by two subsystems: (1) a gyro vertical reference unit

mounted in the wheel house; and (2) pendulum-type sensors mounted at the bow the EMI wave gauge housing. Comparison time series and spectra are included forthese two sensors at several

pitch and roll magnitudes, representative of the range of values encountered during the sea

trials.

The data presented in Table i and in Figures 7 through 11, summarizes the gyro/pendulum

sensor comparisons, Table i is a tabulation of the roll and pitch amplitudes and periods for five data records. In Figures 7 through li, a 128 second time series of the corresponding data is plotted along with the associated frequency spectra (Hz).

(24)

Page 17

TABLE i

GYRO/PENDULUM PITCH ROLL SENSOR COMPARISON

RMS Double Significant

Amplitude Period Data File ID Reference Data

Sensor (degrees) (seconds) Plots

I

Roll gyro 0.6 15.7

Roll pend 0.6 4.7 131800

Pitch gyro 0.2 10.2 10/25/88 Figure 7 Pitch pend 0.1 3.9 13:21:51

Bow accel <0.05 g 7.2

Roll gyro 4.0 19.7

Roll pend 4.2 18.5 16220

Pitch gyro 0.5 16.0 10/23/88 Figure 8

Pitch pend

04

7.9 16:22:34

Bow accel 0.05 g 10.5

Roll gyro 1.5 15.3

Roll pend 3.4 9.3 083801

Pitch pend 1.4 7.1 14:54:47

Roll gyro 2.3 15.6

Roll pend 4.3 8.5 @l6HzO

Pitch gyro 1.1 8.0 10/20/88 Figure lO Pitch pend 2.3 7.1 14:54:47

Roll gyro 1.3 15.5

Roll pend 2.6 8.8 17550

Pitch pend 2.1 8.3 17:55:31

(25)

12.hh

Pitch Gijro

-L 500

4.000

-kolI Pe!

-2.080

h GYRO h168 P tch +h.888 Page 18 I I I I 16 32 48 64 88 !ower Spectre Hs e.2 T- 1.2 PENDULUM 3,543 II Figure 7 GYRO/PENIIJLIM COMPARISON File: 131800.DAT Time: 25 OCT 88 - 13:21:51 ls .3 T

19

8.20 oer Spectrum !ower Spectruffi s 1.2 ts 4,7 848 128 96 112

(26)

13.880

_{Pitch P*nd.}

i-11.880

-1.880

9188 r4J

PetiL

-3.000

6.080 -6.008 GYRO 3.848 +0.000 869 +8.888 0.88 O üwer Spectrum 1s 1.fl rs- 16.8 !ower Spectrum = 8.1 ts 19.? +8.888 0.20 0.48 _0.08 Figure 8 GYRO/PENDULUM DATA File: 16220.DAT Time: 23 OCT 88 - 16:22:34 +0. 800 379.636 rower Spectrum 4s 8.? Ts 7.9 !ower Spectrum ils: U. ts 18.5 0,20 0.40 Page 19 8 :2 64 96 128 168 192 224

26

PENDULUM 8.775 Pitch

(27)

9hh

_{-PcIl PenJ.}

-lîøh

-5îh

GYRO 4.459

Roll Cji'o

Pitch j j rower Spectrum hz 2.9

ts

15.3 8.2h !oieP SpectriM lsr 1.4

T-

8.1 128 PENDU LU M Z4.2h5 Pitch 113.763 16h

+hi

192 224 !ower Spectruol 2.7 Th- 7.1 !owcr Spectrum 1s 6.?

ts

9.3 8.48 256 Figure 9 GYRO/PENDULUM DATA File: 083801.DAT

Page 20

_{Time: 20 OCT 88 - 09:15:10}

15.hhh

_{Pih P*nd.}

(28)

i-t.8

2.888

4.888

18.888

5.808

18.888

5.880

GYRO 8.953 51.368

itch Gjro

I(OIL Fe!d.

RolI G9ro 8 Pitch Rodi 16 8.66 8.28 32 rower Spectrum Pis 2.2 81 Fower Spectrum $5:: 4,5 tS:: 15.6 846 48 64 +6.886 113. 165 08 PENDULUM 39.858 Pitch Figure 10 GYRO/PENDULUM DATA File: @l6HzO.DAT

Time: 20 OCT 88 - 14:54:47

96 8.26 112 rower Spectrum Is 4.6

T-

7.1 Power Spectrum Pis U.S tsz 8.5 8.18 128 Page 21

(29)

16.88 Pitch PnJ.

888 5.888

5.888

25.151 +8.888 8.88

-3808

8.880

1k1I Pei

PLtch 8.28 owev Spectrum ls 2.8 Th- 9.3 rower Spectru 2.5 15.5 0.48 GYRO PENDULUM 81.325 Pttc)i +8.886 187.824 +8.888 0.00 8.28 Fower Spectru t= 4.2 T 8.3 rower Spectre u 5,1 r 0.40 Figure 11 GYRO/PENDULUM COMPARISON File: 17550.DAT

(30)

As shown by this table and Figures 7 through li, there are several differences between the gyro and 'endulum sensor data:

For pitch angles above about 1 degree, both the pitch and roll pendulum RMS double amplitudes are higher (by nearly a factor of two) than the gyro amplitudes.

The pendulum sensor data (especially roll) generally contains more high frequency energy (and thus has lower significant periods) than does the gyro data.

The presented data indicate that the pendulum sensors were responding to translational

accelerations as well as the angular vessel attitude. The pitch spectra for both the gyro and pendulum sensors have "shapes" that are similar to each other as well as to the acceleration

spectrum; spectral peaks among these sensors are well aligned on the frequency axis. Spectral

power, however, is significantly higher for the pendulum sensor, which also tends to have

somewhat more energy at higher frequencies. The additional energy in the pendulum spectrum appears to be due to the vertical (heave) acceleration which adds energy at essentially the same

frequencies as the true pitch spectrum.

The data for the pendulum roll sensor clearly shows the vertical acceleration coupling. Comparison of the gyro roll, pendulum roll and acceleration sensor spectra shows that the acceleration energy has, in effect, been "added" to the pendulum roll spectrum. The higher

frequency content of the acceleration-induced "roll" results in the much lower significant periods of the pendulum roll versus the more accurate gyro roll. The data in Figure 10 show this effect particularly clearly. The comparison of measured data demonstrates that the gyro sensors

measure the angular attitude of the vessel more accurately than do the pendulum sensors even at relatively low accelerations.

(31)

Figures 12 and 13 show the corrected _{wave heights, vertical acceleration and vertical bow}

displacement obtained using the pendulum angles and gyro angles respectively for geometric correction. These figures include a 256 second time series and the corresponding spectra for each parameter. For the low range of pitch and roll values encountered during the sea trials, the differences are small. However, for higher sea states with correspondingly higher accelerations and larger pitch and roll angle errors, the gyro sensors are clearly preferable.

5.2 Vertical Acceleration Sensor Subsystem Performance

Vertical acceleration was measured by a Schaevitz Vertical Accelerometer. The vertical bow displacement was computed from the vertical acceleration and gyro pitch and roll data as

described below:

1. _{Compute the true vertical component (Aa) of the measured vertical acceleration}

(At)

where:

Page 24

define:

rl = sin 6 cos y cos ß + sin 6 sin ß

r2 = -sin 6 cos y sin ß + sin 6 cos ß

ql = -cos y cos 6 sin O

q2 = r1(cosOsin)

q3= r2(cos9cosc)

5 = vertical dip angle (from horizontal) = 78.0 y = sensor azimuth (from forward) = 0.0 deg ß = transverse tilt (from horizontal) = 4. 1 deg

O = pitch angle (positive when bow is up)

= roll angle (positive when starboard side is down)

(32)

2.273 Pi

-l.63

e. ig

-8. 198 18654 -Bo Di -16.293 9.581 -9.861 8.256 963.114 h 48545e Corr. +8.060 e. ee Cori'. Wawe Ht. 8 32 64 96 128 168 192 224 256 Bow t»

Figure 12

PARAMETER DATA USING PENDULUM SENSORS

File: 080301.DAT

Time: 20 OCT 88 - 09:15:10

rouer Spectret Is 0.3

Is

7,4 Power Spectrum Hg: _16.3 Ti:: 8.?

Page 25

24,285 Pitch fower Spectrun

p1s 2.7

Ir

7.1 ve Ht. rower Spectrue Hi- 14.1 Te:

88

6.20 0.48 8.60 0.80 1.86

(33)

1.162 P1ch I f.841 0. 190 -9.912 4.459 0.256 +0. 800 903.860 DI +0.000 9.557 Corr. We e H P I tc h 32 64 96 128 J 192 224 256

!oer Spectrunl

1.4 1g 8.1 Power Spectr 0.3 T8= 7.4 Power Spectru Bg:: _16.3

1;

8.? Figure 13

PARAMETER DATA USING GYRO SENSORS File: 083801.DAT

Page 26

_{Time: 20 OCT 88 - 09:15:10}

825.735

Corr. ti'e Ht.

SpectruM

B. 17.3

8.6 0.80 1.80

(34)

Remove the trend from the time series.

Band pass filter the series in the frequency domain by applying a rectangular window (perfect filter) to the FFT coefficients. The window (W) has W(f)

i for frequency (f) between fi and f2; W(f) = O for all other frequencies. For

the data presented in this report, fi = 2 Hz and f2 30 Hz.

Inverse-FFT the filtered FFT back to the time domain; then integrate the filtered series using trapezoidal integration to compute the velocity time series.

Repeat steps (2) through (4) on the velocity data; compute the vertical bow

displacement time series = g * (integrated velocity), where g = gravity

acceleration.

The acceleration measurement and vertical bow displacement computation functioned well for

the sea trial data. Integration accuracy on the order of i or 2 percent was obtained using

calibrated inputs at the frequencies of interest. Additionally, comparison of the acceleration time series and bow displacement time series as illustrated in Figure 13 shows the expected phase reversal but similar shape of the two traces with no indication of integration artifacts such as excessive low frequency modulation or drift. Finally, using the acceleration significant periods from the acceleration spectrum plots and the amplitudes on the time series plots, the amplitudes on the bow displacement plots are found to be in good agreement with the expected values derived from the approximation:

bow displacement = 1/2 g (acceleration) (period/4)2

5.3 Wave Surface Range Sensor Performance Evaluation

The slant distance range to the wave surface was measured by an EMI infrared wave gauge. This sensor measures the time for an infrared light pulse to be reflected from the sea surface Page 27

(35)

back to the sensor. By transmitting a series of pulses, a continuous measure of the distance to the wave surface is obtained. Electronic filters and signal processing circuits integral to the EMI sensor minimizes the effect of spurious reflections from sunlight, rain and spray (Ref 6). The optical system consists of a concentric transmitter/receiver and the transmitting lens collimates

the emitted radiation to a 0.6 degree beam width. At a range of 75 feet this results in a 1 ft

diameter illumination spot.

The EMI laser sensor operated without problem during the sea trial. On the eastwardcrossing, it survived "green water" submersion and the associated wave impact loads, thus demonstrating its robustness. However, many of the EMI wave gauge data records contain periods of what are termed "data drop outs" - periods of i to 3 seconds when the wave surface range values reported by the sensor remain essentially constant. These data drop outs distort both the raw range data statistics as well as the encountered wave height spectra and statistics.

Three questions relative to the wave range data are discussed below:

What is the envelope of pitch conditions within which the EMT sensor data is acceptable or recoverable?

What is the result of the distorted wave surface range time series on the calculated wave height spectrum and statistics?

Can the distorted data be sufficiently recovered to provide usable wave height information?

To answer these questions, approximately 20 percent of the data records, representing the full range of sea states encountered during the sea trials, were re-processed to: (I) compute drop out statistics as a basis for comparison; (2) relate sea state conditions to drop out severity; and (3) evaluate the effectiveness of algorithms which attempt to reconstruct the wave range data during drop outs.

(36)

5.3. 1 Data Drop Out Detection Algorithm

Referring to Figure 14, the wave surface slant range data record is scanned and each point is compared with the preceding point. If the difference in values is less than a preset limit

(currently0.5 feet), a "sequential flat point" counter is incremented. If the counter reaches one second or more, AND the difference between the last "flat" value and the first "flat" value is still less than the point-to-point difference limit (0.5 feet), a potential drop out is flagged in the algorithm logic. A "true" drop out is detected if the potential drop out is followed within a

preset time by local maximum or minimum of sufficient "height" above or below the drop out.

Details of the drop out detection and range data reconstruction algorithm are presented in

Appendix C.

This peak detection algorithm is essentially empirical and to some extent arbitrary. _{It evolved} by means of trial and error in an attempt to duplicate the drop out "detection" performed by eye, but with a consistet definition which could be applied to all data records. Comparison of the raw uncorrected EMI wave gauge data against the reconstructed EMI data in Figures 15 through 18 indicates that the algorithm is generally in good agreement with visual judgements. It should be noted, however, that the quantitative results, in terms of number of drop-outs and drop out

percentage are sensitive to the threshold parameters used by the algorithm. The "drop out

percentage" can vary by 10-20% depending on the choice of threshold values. Nonetheless, the

algorithm was applied consistently to the data presented in this report, and the results are

considered at least qualitatively correct.

(37)

Page 30

Start mid-\_J pomt Parabolic interpolation

/

m Width

-> Threshold

0

Peak following flat

Start

mid-point

Parabolic interpolation

Parabola fit through points

₍

_{for upward peak}

Parabola fit through points Ø ( _{for downward peak}

Figure 14

Flat Spot Detection/Reconstruction

® Peak following

- flat spot

Heavy line = measured data Dotted line = reconstructed range

(38)

-1489

58 555 +49,824 59.318

8619

P1 cl

p'

I ch Gro ki V _y ci G910 p' w V Ii' 2 64 96 128 168 192 224 256 U N CORR E CTED RECONSTRUCTED I I I I I I 32 64 96 128 168 192 224 256

Figure 15

PECH AND EMI RANGE

File: 083801.DAT

Time: 20 OCT 88 - 09:15:10

It) 1 I Page 31

(39)

-1.658 65.684 +39.898 69.858 +34,711 0.999 0.999 Pitch Cvo -1.658 I Rane I

ji

I T T I I 16 32 48 64 88 96 112 128 16 32 18 64 88 96 112 128 UNCORRECTED RECONSTRUCTED Figure 16

PITCH AND EM! RANGE File: @i6HzO.DAT

Page 32 _{Time; 20 OCT 88 - 14:54:47}

'1 Pitch Cero

(40)

Pitch ro -1h39

-1639

64.643 +42.775

,

f

tr\ ¡

\

! V 1j £4.643 I Rane

ri

I

¿dl

V

t366

PitUi CPQ EflI EdnÌe A t J t ( 96 128 168 192 224 256 UNCORRECTED Figure 17

PITCH AND EMI RANGE File: 12590.DAT Time: 22 OCT 88

_{- 12:59:24}

32 64 96 128 168 192 224 256 RECONSTRUCTED

Page 33

32 64 +42.775 0

(41)

r.ziz

rpitch Cero 68877 583 e 212 -8.782 68.675 +49.583

i V"Î\j' V

EVIl Iane 8 1I Raue I I I i 16 32 48 64 88 96 112 128

)

UNCORRECTED V

il'

r i 16 32 48 64 80 R E CON ST R U C T E D j

t

96 112 128 Figure 18

PITCH AND EMI RANGE File: 134200.DAT

(42)

Table 2 and Figure 19 present the results of the drop out analysis of 15 data records

representative of the range of sea state conditions encountered during the sea trials. The data are presented in two segments: records with pitch periods less than 9.5 seconds; and data with pitch periods greater than 9.5 seconds. Figures 20 through 27 present the data, in order of increasing drop out severity, for pitch periods under 9.5 seconds; Figures 28 through 34 present the data for periods greater than 10 seconds. The wave height time series and spectra in all figures represent "drop out reconstructed data". Each of these figures show the raw wave range time series and spectrum; the "drop out reconstructed" range time series and spectrum; and time series and spectra for gyro pitch, vertical acceleration, bow displacement and wave height. Figures 35 to 38 present, selected time series and spectra derived from the as-measured (non "reconstructed") data. These plots correspond to the plots of reconstructed data presented in

Figures 20, 30, 32 and 38. All angle data used in the processing of the presented data was

measured by the gyro sensors.

Within each segmtit, the Table 2 data entries are ordered by increasing "drop Out percentage", defined as the ratio of the number of points in all detected data drop outs compared to the total number of points (2048) in the data record. The table also lists the number of different data drop out (#) segments in the record, and the gyro pitch double amplitude and period. The filename of the data record and the corresponding data figures are included for reference.

Although there is a fair amount of scatter in the drop out statistics due to the sensitivity to the threshold used in the algorithm, a qualitative relationship of the drop out severity to the sea state,

specifically to the pitch amplitude and period, is evident. Drop out severity is generally

proportional to pitch amplitude. To a lesser degree, the number of drop outs is also a function of pitch period. The drop out percentage at lower periods (higher frequency) is relatively greater than for those same pitch amplitudes at longer periods. Figure 19, based on the data in Table 2, clearly shows this dependence.

(43)

Examination of the unprocessed wave range data in Figures 20 through 34 shows that a

qualitative difference exists between the drop outs at lower pitch periods compared with higher periods: _{At the lower periods, the wave range data drop outs occur shortly after both pitch} minimums (bow down) as well as pitch maximums (bow up); for the higher periods the data drop outs occur only after pitch minimums.

Page 36

TABLE 2

DROP OUT DETECTION STATISTICS

Filename RMS Pitch

Double Significant Dropouts Reference

Amplitude Pitch Figure #

(deg) (seconds) 131800 0.1 10.2 3 2.6

20&35

16220 0.5 16.0 13 _7.6 21 13110 0.4 16.8 4 7.9 22 165800 0.4 14.9 8 8.6 23 17480 0.2 11.7 18 10.2 24 12050 0.8 13.4 17 12.1 25 12590 0.6 16.4 9 14.8 26 17180 0.1 14.6 21 16.6 27 134200 0.3 8.7 9

l2J

28 14030 0.7 8.8 25 20.5 29 083801 0.7 8.1 29 25.7

30&36

08220 1.0 8.3 21 27.1 31 @l6HzO 1.1 8.0 22 34.1 32 & 37 08380 0.8 8.5 38 46.1 33 17550 1.4 9.3 43 56.1

34&38

(44)

(Numbers in parentheses are significant pitch periods) (11.7) :: (8.7) (10.2) (8.3) (14.9) (16.4) ç> (14.6) (13.4) (16.8) ç> (9.3) F L A T

_S

p

o

T S

4d.iF1Ù

2MMHII

(16.0) (8.5) (8.1) K(8.8)

ti

tiItiui

ti.4tiEitlIl

t1.8E1F1ti

i

.

2iE1ti

i

.(I1iii*1

}MS CYJ«:p PiTCH DOUILE AMPLiTUDE

Figure 19

(45)

-t 224

se.1s4

PITCH AND UNCORRECTED EMI RANGE

324

M

_J

\t

\/kJ

- 491 1J hne S. 122 -S 125 I. 123 -1524 t.551 - IIUI. -4. 535 16 -3.57? 1 fl

PARAMETER TIME SERIES

13. 381

'k

t.tee t.tP t 21 S. NS s.. -1. 125 +I.IN 112 128 I.N II k 1S.U7 I hate

.I

ttd s.. S.N

UNCORRECTED EMI RANGE SPECTRUM

kw tecal

C.er. u.u. at.

A S. t 41 PARAMETER SPECTRA Power Spectru. St 4.9 Is- 3.2 Pu.sr Ipectr 3.3 1.N Pawsr 1psctr. 55r 5.2 Ts 11.2 $ai.r p.ctr Sir 5.5 T.. 7.2 Pius, p.ctr.. t..

lii

P...r *..ct. S.. E.2 t.- 3.1

r

-SN LS

PITCH AND RECONSTRUCTED EMI RANGE RECONSTRUCTED EMI RANGE SPECTRUM

Figure 20

SEA ERIAL DATA

File: 131800.DAT

Page 38 _{Time: 25 OCT 88 - 13:21:51}

+51 115

I 16 32 41 64

121 32 II 64 Bl 96 112

(46)

0.437 -0. 856 63.201 +45. 338 63.311 +45 331 -0.65$ 0.835 Iii Pitch Ç.PC 11 hnù 5.437 PIteh _$i.s

I

su. I. 523 +5. 2ES. 135 +1. IM S.N

UNCORRECTED EMI RAN(F PECTRUM

553.63e Li h.,. IM 124.704 Cr. .w. St. +,. s. u.s PARAMETER SPECTRA Pewsr psctrU is 11.8 15.4 l.N cis 1p.ct,. Us 1.1 tuc 15.5 Pm.p Sp.cti. k 1.1 r. 11.5 P...Spucir.. b I.? tS 16,5 Pur Ipectru. I. 7.2 ti- 11.1

I..

i.N 125 151 12 U4 255

r

i

256 0 32 64 96 128 160 192 224

PiTCH AND RECONSTRUCTED EMI RANGE _{RECONSTRUCTED EMI RANGE SPECTRUM}

-I $41

5. 354

.5 SII

.-a. lupi.

Figure 21 SEA TRIAL DATA

File: 16220.DAT

(47)

B 549 Pitch CBi'o B 54? 62 166 62.151 1.537 PItc -1515 411 1l B.ne

/

'46.86! r 32 64 96 128 161 192 224 256 Iii liaSe

PITCH AND RECONSTRUCTED EMI RANGE

-I. 153

4.171

-Is IIi.

4.I

tørr. Ma Ht.

+1.

121 I» I.1

RECONSTRUCTED EMI RANGE SPECTRUM 2.561

+1.

1. IN

ti. _

B Il 6 26 6 46 6 66

PARAMETER TIME SERIES _{PARAMETER SPECTRA}

Figure 22

SEA TRIAL DATA

Fije: 13110.DAT

Page 40

_{Time: 23 OCT 88} _{- 13:11:41}

B BI _{i ee}

Ill

P.ur Ip.cti. ou u.. T. 16.1 PaNI, Bp.etr'.0 r 1.1 ?s

III

!hh11P Sp.ctp

l7.2

ta 26.4 Pa.er 2p.ctr k. 7.3 i. 13,6 PITCH AND UNCORRECTED EMI RANGE _{UNCORRECTED EMI RANGE SPECTRUM} .54 PItc CVo i) -1.547 H % 121 1I 12 224 Z5 I 32 64 iii

ii

_Ifl 224 256 l.N 1.21 IN 6.61 I. N N

(48)

0.427 -0.432 1.124 -1141 3.11 Sd ILUpI. 192 224 256 PITCH AND UNCORRECTED EM! RANGE

I 32 S 121 161 192 224

213.458 ia a.

0.00 1.2$ 0.40 0.61 I.N

RECONSTRUCTED EMI RANGE SPECTRUM UNCORRECTED EM! RANGE SPECTRUM

1.41 0.61

Pawrn Spsctr hr 71 tir 11.0 ?a.,r Ljuectr.. hr LI ti; 12.? Sp.ctr I. S.l t.r 14.9 Pm.r Sp.etrU U.:

Ii

T. 10,8 Spiet.. hr S.! t, 14.4 P..r Op.ct... Figure 23

SEA TRIAL DATA

File: 165800.DAT

Time: 23 OCT 88 - 17:03:47 _{Page 41}

0.20 0.40 168 1.324 PItcb 3PII -0.547 0.324 PItc g,c

t

-L 547 ?. 9.0

I..

1.00

(49)

1143 PItctt Curo Pt '1 î .

\1\J

]

1Pj\

kd, W"I! _j, -LS" -e 1.143 PItcbGar. -"Sl, 51. +5!. 545 I. 333 I 32 14 % l 1M 12 224 256 64 96 l2 16! 192

Page 42

I ii À1 e.esr 224 256 I 14 5.2! 13.161 0.

t..

-1.323 U.N Figure 24 SEA TRIAL DATA

File: 17480.DAT

Time: 24 OCT 88 - 17:48:26

.

Mt.

s..

Zp.cti is 1.4 Tir II.? pctr.. Si: 3. t.& 11,5 Pier Upuctr S.S ts 5.2 U.N 1.11 S, 555 '1 hne

I.N

ia

1.41 SII _5.N

i..

Power Spectruo

Ir

c ₄

SS

33,374 Iaoe

PITCH AND UNCORRECTED EMI RANGE _{UNCORRECTED EM! RANGE SPECTRUM}

PITCH AND RECONSTRUCTED EM! RANGE _{RECONSTRUCTED EMI RANGE SPECTRUM}

Pow. p.ct,.

Su: 1.1

Tir I.?

(50)

61.759

+49 823

i hne J

123

6. S22

PITCH AND '!N'ORRECTED EMI RANGE

-PI

1h 165 192 4 2S6

14541? kiie

S, see

e es 8.28 5.48 5 68 8 85

UNCORRE'TED EMI RANGE SPECTRUM

IN . 513

5.71.3 PIt _{Pller lHttIUU}

Tr 13.4 k 1.6 il. W I. 142 +5. -465 156 +1. Aceál. lupi. A I.1 I.N PARAMETER SPECTRA Pe Spectr..

l,

95 T 11 3 I Il

Poiler Spec trui

r

9.5

tr

11.2 1,M Pail., p.etrM

jr

1.2 Tri 7,9 Pail., p.etr..

ir

11.9 tri 11.7 Paw., p.ctr Sri 9.8 tri 9.6 5.65 I.N IN PITCH AND RECONSTRUCTED EAI RANGE _{RECONSTRUCTED EMI RANGE SPECTRUM}

Figure 25

SEA TRIAL DATA

File: 12050.DAT

Time: 21 OCT 88 - 12:05:20 _{Page 43}

I '? PtcIi iro -1 278 256 32 64 96 128 165 192 224 8.732 'PIte pc -1.271 32

s

192 324 256

i.

--s.

(51)

e 3â rrItc CIPO 64 643 i42.775 t.%O pita cp. 64.43 .42.7Th -15% 7.22$ -o173 7.743

-31

hije

y

I 32 44 % 12$ ia I2 324 PITCH AND RECONSTRUCTED EMI RANGE

Page 44

e.ree -

-32 64 96 128 161 192 224 256 e ee e.z e41 8.68 8 1 88

488 845 :

fi. fis.

e.0

RECONSTRUCTED EMI RANGE SPECTRUM

L

Figure 26

SEA TRIAL DATA

File: 12590.DAT Time: 22 OCT 88 - 12:59:24

i."

PARAMETER SPECTRA e.0 $s 13.2 T 11.1 PSZ Spuctrà k. ti. 11.4

lu

I. 15.2 t,.. 1.I Puupur Ip.ctr is. 13.1 ti. II,? IN 11.5% $1. -I. $2$ 573.341 Pl$: !uwsr 2p.ctr jir 1.2 7ur 15.4

kw Accu, sur Sp.etpei

h S.j

T.. 11.1

- Clip1. !' Sp.ct

(52)

t. 971

-1.321

62.561

+44 457

I 32 64 96 128 16! 192 224 256

PITCH AND UNCORRECTED EMI RANGE 1.371 PIt. Gpr. 3. ? 1.212 7.432 -il. 326 Pitch 6ro 11 Eane

Ill 111 flZ 324 249.793 24.2V r I.N I II 1.2! 1.41 1.68 8 88

U.N 1.21 U.0 1.61

161.443 _CltrlMv.Mt.

PARAMETER SPECTRA Puwsr ip.ctr is 12.3 ts 12.5 mr Zp.et PUSP Spictr, kr 12.7 i. 13.7 P.w.r Ipsctr k. 11.1 I. 1.61 S.N IN 11.5

III

1.N

I2.1

t. 14.6

b

1.1 T. 3.3 Figure 27 SEA TRIAL DATA

File: 17180.DAT

Time: 21 OCT 88 _17:18:20

(53)

?1Z Pttci Curo -L712 1.472 PItc 68.877 hui 8 16 32 1.212 PItck 6po I 11 32 II 14 N % 112 12! k 1 '

VV"

I 16 32 41 4 N S 112 121 112 121 8 88 8.28 8.4! 0.6! 51.188 r I.88 U.418 47.114

f..-I.N 1h laut Isw lessi lupi. 8.41 1.11

PARAMETER TIME SERIES PARAMETER SPECTRA

Figure 28

SEA TRIAL DATA

File: 134200.DAT

Page 46

Time: 24 OCT 88 - 13:44:43

88! 1.08

Puwsr Sptr lue 7.1

t, 7.4

u.0 u.0 u.0

Puwur tp.ctr ?.. 1.7 Pswsr £p.etP h: 8.1 ?u 6. Puiiur $psctr tI ,1 PITCH AND RECONSTRUCTED EMI RANGE RECONSTRUCTED EMI RANGE SPECTRUM

PITCH AND UNCORRECTED EMI RANGE UNCORRECTED EMI RANGE SPECTRUM

-8. 484 I. 872 $sw *ccI. Power Spcttr. 36. 262 0c 7,1 T: 71 w 'J y' Pais, Ipsctr b. 6.2 t.. 7.8 u.N

(54)

62 800 r +42.054 32 64 96 128 160 192 224 -1,41 .188 w

11! I.$e -1.114 8.502 - lupi.

f

256 0.00 0.20 0.40 0.68 0.80 1.80 +1. I. PItc $1. 733.111 Ii i. s.. 477.511 tDpp. I. N

8e

Nt.

s..

PARAMETER SPECTRA Powir Spectr iu- 15 4 Ti' 6.? 7.6

!"

lp.ctI,i Sr 3 Ti S.l Paw., lp.ctr.I Ur 0.3 T.. 7.5 ap lp.ctr I.. 16.5 t.. o.' Spsctr I.. 15.3 t.. 1.61 0.11 1.11 Figure 29

SEA TRIAL DATA

File: 14030.DAT

Time: 20 OCT 88 - 14:03:55

_{Page 47}

718

64 % 120 la 102 UI 256

PITCH AND RECONSTRUCTED EMI RANGE _{RECONSTRUCTED EMI RANGE SPECTRUM} cI

0. 324

64 ₁₁ 102 4 286

(55)

1.11, PI¼I +13,124 I.31, P1 51.311 +4e. 147 t. 19$ mi k* -9.112 64 % Ifl III 64

'ç

% 2.21 III 132 112' l 256 't

1.162 .-1I -1.141 4 256

J

224 25 51141 4I, pl

U.2I

iii

Ill

IN IN

.21

Figure 30

SEA TRIAL DATA

File: 083801.DAT

Page 48

_{Time: 20 OCT 88 - 09:15:10}

Puwsr p.ctrd Ii: 6.5 ti' 7.1 Zpsrti.a 1.4 tm V.1 Piip 3ip,ictN k 13 74 L t ru. ; 16 3 O? ! *.ctrs. Is 141 'tas

li

U.N 1.11 PITCH AND UNCORRECTED EMI RANGE _{UNCORRECTED EMI RANGE SPECTRUM}

(56)

5.51 +43 11 1.732 s,.s1 442.V 1.137 I. 231 21. I I 32 64 % 121 1M 12 324

I 32 4 % 121 1M 12 324 6

M

J I

I31,51

ri ii,.

+1.

+4. 1122.161

Pijas, p.cta k: 5.4 Ti. 7.7 I PARAMETER SPECTRA P..sr Ip,ctr iv 11.? ti. I.3 !' Ipsetim ii: 21.4 ti. 5.4 Figure 31 SEA TRIAL DATA

File: 08220.DAT

Time: 19 OCT 88 - 08:22:52

_{Page 49}

5.21 s..

I..

IN

$p.ctrs

I. 2.5

!' .3

1.174 PIteE' I 64 In lu 112 214 112.112 lit. P.m.., Ip.ctrs. 1. n.3 ti. 156 5.61 U.N i.N

(57)

I tItd 4PC

S. l4

s ,, Pite Cp

-1 SI

I' 32 4 N S 112

1111.54?

4I.NIr

-I.N P.11 I. I.0 I.N

p ita 5cm k. 273 t.. 9.7 IN !u11.r Ipsvti. is 2.2 r.. s P..P 1p.ctru Ps.,r Ipsctr 5.. 21.3 t. 7.7 l.N Figure 32

SEA TRIAL DATA

File: @l6HzO.DAT

Page 50

_{Time: 20 OCT 88 - 14:54:47} I _+34.711 I 32 S ₁₁₂ ₁₂₁ 7J usr peti. S.: Us T.. 74

(58)

676 rl? 3(1 -1.2s' I. 939 -I 876 U.S3 Pitch pc 8 32 64 96

PITCH AND UNCORRECTED EM1 RANGE

Pl tc -e isa 13.977 _a _'lupi, 54 96 L g

4V!

128 168 192 224 p

I

Iu

2S6

fr

I ji

192 224 2S5

"w !'wsr Ip.ctrU I. 1.5 Î, S5 i.p Sp.ctIM br 1.3 ,r 7,fl !'wIP Sp.ctr.. , 19.I ?. 9.8 !w.P $p.Ct. 1. 16.9 18,3 U.N Figure 33 SEA TRIAL DATA

File: 08380.DAT

Time: 20 OCT 88 - 08:38:01

_{Page 51}

Pitch Cero I 563 V -1 25k ¿4 X 121 lU 192 224 2S I 32

(59)

I 322 TPitcI ro S9. 83? +29.958 +24.!14 -1 548 U. 285 8 32 64 96 121 168 192 224 256

'E

I 32 M t 181 182 324 S

748 634

+8.

U.N

Ita

I 11t$ _{?ower Spettri.}

sa

1.51 PARAMETER SPECTRA r

--Ts- 9.4 !'' p,CtP. I. 5.4 Ti 1.3 Spsctr P,.sr 5pìctr r 28.4 t.- 8.8 p., 2p.ctr i.. 25.8 t.- 11.2 U.N IN Figure 34

SEA TRIAL DATA

File: 17550.DAT

Page 52

_{Time: 19 OCT 88 - 17:55:31}

888 1,88

8.88 I 28 8. I 8.68

(60)

3.122 -8.824

L89 -Row DIp1

e.889 3.169 rCor

-4.

8.868 +8.88e PItch IJáve Ht. Row flcce!.

__pJ\J

-+8.888 8.888

Dicp.

15.2S1 _{Corr. Uave Ht.} 6 ₃₂ ₄₈ ₆ 8.48 I J I 88 96 112 128 8.68 8.88 o,er Spectrun 8.2 T _18.2 f ow er

N:

8.8 Tsz _?.2 Spectru Spec&ria Power _Spectrvii

1M

Figure 35

(NONFLAT SPOT

_{RECONSTRUCTED)}

File: 131800.DAT

Time: 25 OCT 88

_{- 13:21:51}

Page 53

8.28 +8.888 e. me

(61)

1.162 PL -lI. 267 9.851 -14.65? 4.459 L. 256 ch I -L. 199 11.716 -fio Divpl. P I t ch

fleme

913.868 -h Di

$g 16.3 Tt= 8.? 2 64 96 128 161 192 224 256 !o1er Spectrufl Ms 1.4

lÌ-

8.1 Spec ti't 1.3 Ts 7,4 Spectrum Figure 36

(NONFLAT SPOT RECONSTRUCTED) File: 083801.DAT

Page 54

Time: 20 OCT 88 - 09:15:10

825.735 Corr. ve Ht. rower Spectrmm

0* 17.3

+0.101

PL.

t8

8.6

(62)

0.239 -8 257 17. 148 -13.756 16. 671 -25.745 8.953

+88M

0.494 1415.414 -C +0.888 8.08 P1tc1 16 +8. 088 1312.293 -Bow Dutt. vr. Wáve H 0.20 32 -48 T 64 0.40 0.60 88 96 Figure 37

(NONFLAT SPOT

_{RECONSTRUCTED)}

File: @l6HzO.DAT

Time: 20 OCT 88 - 14:54:47

112 128 rower Spectrim ls 2.2 It- 8.8 !ower Spectr*w Bg:: _27.3 Ti= 9.? !ower Spectru 8s 31.6 TB 7.6 8.80 1.00 Page 55 Bo Aec !°'" Spectrfl g _8.5 I's 7,4

(63)

1.519 -1.548 8.295

-02m

15.976 -Dow Diipi. -16.929

/

18. 661 -34.189 60.657 8.958 Pitch Dow Accel. +0.080 6287.888 now O 18895.698 Corr. +0.080

-0.88

Uve Ht.

ML

8.28 8.48 8.68 1' !owev Spectruni 2.8 9.3 rower Spectrs 8.4 Ts

93

!oweP Spectre. Di: 29.4

ti

9.8

!ower Spectru

H. 48.8 te:: 9.6 8.88 1.80

Figure 38

PARAMETER TIME SERIES (NON FLAT SPOT RECONSTRUCTED)

File: 17550.DAT

Page 56

_{Time: 19 OCT 88 - 17:55:31}

1.

(64)

Three possibilities have been considered for the drop out behavior of the EMI sensor: (1)

infrad beam occlusion (absorption or dispersion) by "sheets of water" splashed up when the bow ..tarts down from its uppermost position, and/or thrown up by the bulbous bow of the SL-7 when the bow starts to move upward from its lowermost position; (2) rates of change of wave

range greater than the rate of change threshold (20 centimeters in 50 milliseconds or 13.1

feet/second) set within the EMI sensor electronics; and (3) a malfunction in the EMI sensor. Of the first two possibilities, beam occlusion is considered more likely than range rate threshold exceedance since even for the most extreme pitch/bow movements, the range rate of change (average rate over 1/4 of a cycle) is more than a factor of two below the 13. 1 feet/ second EM! threshold.

The third possibility for the data drop outs must be considered since it can neither be verified nor ruled out by the data collected during the sea trials. It is possible that the EMI wave surface range sensor data validity "decision" logic was malfunctioning or mis-calibrated. At least one experimenter has reported that a malfunctioning EMI wave gauge (also experiencing data drop outs) was returned to the factory and then performed correctly after factory repair which included "optics realignment".

5.3.2 Wave Surface Range Reconstruction Algorithm

In order to assess the impact of the wave range data drop outs on the calculated wave height spectrum and statistics, an algorithm was devised to "reconstruct" the range data during drop outs.

The details of this algorithm are provided in Appendix C; the basic logic of the

reconstruction process was shown in Figure 14 and outlined below.

Referring to Figure 14, the drop-out is assumed to occur near the bottom of an upward

(increasing range) portion of a range cycle, or near the top of a downward (decreasing range) portion of the cycle. The reconstruction algorithm locates the point at the start of the drop out and the point (local maximum or minimum) at the end of the drop out when it is assumed that correct data is again being measured. To reconstruct the range data, the points between the start

(65)

and the midpoint of the drop out are computed by linear interpolation between the range value at the start of the drop out and the range value at the local minimum/maximum after the drop out. The points representing the "peak" of the range cycle (i.e., the pointsbetween the midpoint of the drop out and the following local minimum/maximum) are then computed by parabolic

interpolation using the three points at: (1) the

start of the drop out;

(2) the local

minimum/maximum following the drop out; and (3) the midpoint of the drop out which is

assumed to have the same range value as point (2).

As can be seen by Figures 15 through 18, the reconstruction algorithm appears to perform

qualitatively as expected. Comparing the uncorrected with the reconstructed data, it can be seen

that the reconstructed data has reasonable magnitude and phase relationships with the

non-compromised portions of the range data as well as with the pitch data. Occasionally, the

parabolic interpolation of the peak portion of the range cycle appears too steep, resultingin peak over or under-shoot. This behavior can easily be remedied, for example, by adding some logic to the algorithm to limit the difference between the height of the reconstructed "peak" and the measured following local minimum/maximum.

The non-reconstructed (i.e. as measured) wave surface range data are presented in Figures 35 through 38 for comparison with the corresponding reconstructed data presented earlier in Figures 20, 30, 32 and 34. The significantly lower wave heights computed using thereconstructed wave range data are due to several factors:

Wave range peaks are missed; spurious large wave ìeights result when these

missing range peaks are subtracted from the (in-phase) bow displacement values.

Since they "hold" the range values at the top and bottom of the wave range

cycles, the drop outs cause the wave range to act as though the waves are out of

phase with the bow motion; in effect the wave heights are added to the bow

displacement when they should be subtracted, and vice-versa.