THE NATIONAL --SHIPBUILDINGRESEARCH

PROGRAM

Photogrammetryin Shipbuilding

U.S. Department of Commerce

Maritime Administration

in cooperation with

Todd Shipyards Corporation

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE JUL 1976

2. REPORT TYPE N/A

3. DATES COVERED -

4. TITLE AND SUBTITLE The National Shipbuilding Research Program, Photogrammetry in Shipbuilding

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Naval Surface Warfare Center CD Code 2230 - Design Integration ToolsBuilding 192 Room 128-9500 MacArthur Blvd Betheda, MD 20817-5700

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited

13. SUPPLEMENTARY NOTES

14. ABSTRACT

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT

SAR

18. NUMBEROF PAGES

58

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

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

FOREWORD

This is one of the many projects managed and cost shared by Todd Shipyards Corpora-tion as part of the National Shipbuilding Research Program. The Program is a cooperativeeffort between the Maritime Administration’s Office of Advanced Ship Development andthe U.S. shipbuilding industry.The objective, described by the Ship Production Commit-tee of the Society of Naval Architects and Marine Engineers, emphasizes productivity..-

The suggestion to investigate the effectiveness of photogrammetry in U.S. shipbuildingwas submitted by Mr. Robert W. Schaffran, of the Maritime Administration, whoobtained the idea from foreign literature. Mr. Thomas P. Primavera of the U.S. ArmyTopographic Command contributed the necessary indoctrination in photogrammetrictheory and processes which was the basis for preparing the research specification.

As conventional photogrammetry was applied mostly in topographical work in 1974,the American Society of Photogrammetry advised of only six prospects having actual, orpossibly applicable, experiences in industrial applications. Proposals from the six wereevaluated by photogrammetric authorities from the University of California and theUniversity of Washington. The research work was assigned to:

john f. kenefick Photogrammetric Consultant, Inc.Indialantic, Florida

Mr. L. D. Chirillo, Todd Shipyards Corporation, Seattle Division, was the ProgramManager.

Special appreciation is expressed for the extraordinary assistance, accompanied byindispensable access to real ship design and building projects, furnished by:

Appreciation is expressed to the following for their constructive criticism of the manualin its draft form: Mr. S. A. Mahler of Newport News, Va., a shipbuilding consultant; Mr.G. A. Uberti, National Steel & Shipbuilding Co., Mr. R. C. Moore, Newport NewsShipbuilding; and Mr. P. E. Jaquith, Bath Iron Works.

EXECUTIVE SUMMARY

Photogrammetry is the art, science and technology which includes obtaining reliableinformation about physical objects by measuring and interpreting photographic images.The process began about 135 years ago and has its principal application in aerial surveyingfor compiling topographic maps. But, land-based or non-topographic applications, nearlyas old, are now becoming more numerous because of rapidly occurring technical de-velopments.

Presently, a number of organizations throughout the world productively employ photo-grammetry, on a more or less regular basis, in architectural, biomedical and industrialareas. Within the latter there are shipbuilding applications for the measurement of hullsections, propellers in various stages of fabrication, cargo tanks, etc. Already, and as aresult of initial work on this project, a U.S. shipbuilder has engaged a photogrammetricfirm to prepare sounding tables for spherical LNG tanks.

Thus, this project has not been concerned with whether photogrammetry is a viable andsufficiently accurate measurement technique; it is ! Instead, the main objective has been toinvestigate the applicability of photogrammetry to dimensioning problems which fre-quently occur in shipbuilding. Particular emphasis has been placed upon applicationswhich may have the greatest potential for improved productivity. Time and dollar costsinvolved for four photogrammetric demonstrations, in real shipbuilding situations, arewell documented herein.

Uniquely, the photogrammetric researcher was required to obtain a sufficient under-standing of shipbuilding processes to insure that this manual will be readily understood byshipbuilders. Further, his knowledge of shipbuilding permitted him to report objectively;both advantages and disadvantages are described. Information of immediate interest is inthe first two chapters. Comprehensive appendices are provided for those shipbuildingfunctionaries who will have greater need for understanding photogrammetry.

v

TABLE OF CONTENTS

LIST OF ILLUSTRATIONS

LIST OF TABLES

1. INTRODUCTION

1.1 The Science of Photogrammetry

1.2 The Value of Measuring with Photogrammetry

1.2.1 Advantages

1.2.2 Disadvantages

1.3 A Search for Applications in Shipbuilding

1.4 Summary Opinions

1.4.1 Conclusions

1.4.2 Recommendations

2. DEMONSTRATED APPLICATIONS

2.1 Measurement of a Complete Midbody Section

2.1.1 Preparation

2.1.2 Photographic Procedures

2.1.3 Laboratory Measurements

2.1.4 Data Processing

2.1.5 Evaluation of Results

2.1.6 Time and Cost Analysis

2.1.7 Suggestions Relative to Future Work

2.2 Measurement of a Hold Receiving a Spherical LNG Tank

2.2.1 Preparation

2.2.2 Photographic Procedures

2.2.3 Laboratory Measurements and Data Processing

2.2.4 Evaluation of Results

2.2.5 Time and Cost Analysis

2.2.6 Suggestions Relative to Future Work

2.3 Measurement of a Transverse Bow-Section

2.3.1 Preparation

2.3.2 Photographic Procedures

2.3.3 Laboratory Measurements and Data Processing

ix

x

1

1

1

1

2

2.-

2

2

3

5

5

5

6

6

7

8

8

9

10

10

11

11

13

13

13

13

14

15

15

vii

2.3.4 Evaluation of Results

2.3.5 Time and Cost Analysis

2.4 Piping Drawings from a Machinery Space Model

2.4.1 Preparation

2.4.2 Photographic Procedures

2.4.3 Mapping Procedures

2.4.4 Evaluation of Results

2.4.5 Time and Cost Analysis

2.4.6 Suggestions Relative to Future Work

2.4.7 Long Range Potential

2.5 Summary of Demonstrated Applications

3. A SURVEY OF PHOTOGRAMMETRY IN SHIPBUILDING

3.1 Literature Search and Direct Mail Solicitation

3.2 Review of Suggestions from U.S. Shipbuilders

3.2.1 Applications in Design

3.2.2 Applications in Fabrication

3.2.3 Applications in Erection

3.2.4 Applications in Outfitting

3.2.5 Applications in Sea Trials

3.2.6 Applications in Ship Repair

3.2.7 Miscellaneous Applications

3.3 Productivity of Applications Suggested by U.S. Shipbuilders

Total In-House Capability

Total Subcontract Arrangement

Combined In-House/Subcontract Arrangement

4. OPTIONS FOR APPLYING PHOTOGRAMMETRY IN A SHIPYARD

4.1

4.2

4.3

16

16

17

17

18

18

19

19

22

23

24

25

25

25

. 26

27

27

29

30

31

31

32

33

33

33

33

LIST OF ILLUSTRATIONS

2-1

2-2

2-3

2-4

2-5

2-6

2-7

2-8

2-9

Blowout View of Midbody Showing Locations of Targets

Locations from Which Photographs Were Taken of Midbody

Typical Photograph of Midbody from Ground

Typical Photograph of Midbody from Overhead

Deviation of Midbody Tank Top from Best-Fit Plane

Deviation of Targets Representing Midbody Vertical Centerline

Deviation of Forward Transverse Section of Midbody from Best-Fit Plane

Deviation of After Transverse Section of Midbody from the Plane Parallel to thePlane Best Fit to the Forward Transverse Section

Examples of Best-Fit Circle to Turn-of-Bilge

2-10 Flat-faced Clip for Squaring off Beveled Edges

2-11 Locations of Targets and Camera Stations in an LNG Spherical-Tank Hold

2-12 Typical Photograph from Weather Deck of Hold for a Spherical LGN Tank

2-13 Typical Photograph from Tank Top of Hold for a Spherical LGN Tank

2-14.

2-15

2-16

2-17

2-18

2-19

2-20

2-21

2-22

Deviation of Top of Cylinder from the Best-Fit Plane

Deviation of Weather Deck from Plane Parallel to Plane Best Fit to Top of Cylinder

Deviation of Top, Inside Edge of Cylinder from Best-Fit Circle

Locations of Targets and Camera Stations for Survey of Transverse Bow-Section

Typical Photograph of Transverse Bow-Section

Deviation of Shell Plating of Transverse Bow-Section from Design

Locations from Which Photographs Were Taken of Machinery Space Model

Photograph of Machinery Space Model Showing Extent of Model Mapped

Elevation Drawing of Portion of Machinery Space Model

8

9

10

11

12

12

12

12

13

15

15

16

18

20

21

ix

LIST OF TABLES

2-1 Photogrammetrist’s Man Hours and Cost for Dimensioning the Midbody

2-2 Shipbuilder’s Effort in Support of Dimensioning the Midbody

2-3 Photogrammetrist’s Man Hours and Costs for Dimensioning a Hold to Receivea Spherical LNG Tank

2-4 Shipbuilder’s Effort in Support of Dimensioning a Hold to Receivea Spherical LNG Tank

2-5 Photogrammetrist’s Man Hours and Costs for Dimensioning the TransverseBow-Section

2-6 Projected Man Hours and Cost for Plan and Elevation Piping Drawings

2-7 Summary of Demonstrated Applications

3-1 Applications of Photogrammetry Revealed by Literature Search and Direct MailSolicitation

4-1 Estimated Start-Up Costs for Implementing Photogrammetry In-House

9

10

14

14

17

22

23

1. INTRODUCTION

the object to be measured is large and relativelyinaccessible for conventional measurements,

the object is complex in detail or shape,

a large amount of data is required,

measurements at widely separated locations on theobject must be precisely related to each other, and

the object is in motion.

Because of its potential for handling such conditions,photogrammetry should be considered as being a com-plement to, rather than a substitute for, conventionalmethods of measurement. But, it has an important uniquecapability. Photographs for photogrammetric work arecheaply made permanent records that can be easily filed.They could serve in the future should a need arise toestablish facts, such as for litigation, regarding as-builtdimensions, tank sounding-tables, etc.

Basic methods of photogrammetry are analog, semi-anal ytical and fully analytical. Analog photogrammetryemploys stereo pairs2 of photographs which, when viewedsimultaneously in a two-projector instrument called astereo plotter, permit graphical mapping of the perceivedthree-dimensional optical model of the scene. Semi-analytical photogrammetry is practically identical exceptthat discrete points in the optical model are digitized (X,Y, Z) and then subjected to computer processing for nu-merical and/or graphical presentation of results.

Fully analytical photogrammetry requires measure-ments of images of discrete points directly on the photo-graphs from stereo pairs or on convergent photographs.2

Subsequent numerical triangulation produces a digitalmodel of the scene which can then be manipulated furtherto yield numerical and/or graphical results.

Because of the need to measure the optical model orthe photographs directly, photogrammetry is not a real-time measurement system. Under ordinary conditionsseveral hours to several days are required to produceresults.

1.2 The Value of Measuring with Photogrammetry1.2.1 Advantages

a .

b.

c.

d.

e.

f.

g.

h.

i.

j.

Complexity of shape or detail of the scene isnot restrictive.

Relative directions to all points of interestwithin the field of view of a camera are re-corded instantaneously with a single open-ing of the shutter. (This is in contrast to timeconsuming measurement of horizontal andvertical angles with a transit or theodolite.)

All data (photographs) can be obtained in ashort period, thereby minimizing the effectof thermal induced or other changes withinthe scene upon the consistency of the finalresults.

The amount of data to be extracted from thephotographs does not pose severe time andcost limitations.

The process is virtually noncontacting.

Due to the short period of time required fortaking the photographs, there is minimuminterference with ongoing work. In someinstances work may not be interrupted at all.

Objects in motion can be accommodatedwith the use of two or more cameras havingsynchronized shutters.

The orientation of attitude of the scene is ofno consequence. Photographs can be takenfrom above, below or to the sides so long asthere is room to stand off with the camera.

With the fully analytical method of photo-grammetry, the accuracy can be varied overa wide range to suit a particular need.

The photographs constitute a permanentrecord and may be reused to check resultsor gather additional data.

k. Photographs may be taken and archived forany potential need to produce data in thefuture. This applies to potential litigation aswell as to pure technical requirements.

1.2.2 Disadvantages

a. Results of photogrammetric measurementsare not produced instantaneously. Depend-ing on the difficulty of the task considered,the time required to produce final resultsmay range from several hours as a minimumto several days. An extreme situation mightinvolve several weeks.

b. Useful photographs cannot be taken in rain,fog or snow.

c. Measurements in very confined spaces aregenerally not practical due to the large num-ber of photographs needed to cover the en-tire scene. Also, the depth of focus of thecamera lens may not be able to accommo-date varying distances to points within thescene.

d. The first. cost for an in-house capabilityranges from $50,000 to $180,000 dependingon the capability desired. 1

e. Depending on the capability desired, aphotogrammetrist or special training of ashipbuilder would be required for an in-house capability.

1.3 A Search for Applications in Shipbuilding

An extensive search of photogrammetric literatureand a direct mail solicitation revealed considerable in-terest in applying photogrammetry in shipbuilding or inindustries having similar industrial functions. A number ofsuch articles and reports are described in Chapter 3.However, only three instances were found whereinphotogrammetry is apparently applied as a routine mat-ter. These are: development of sounding tables for LNGspheres, measurement of ships’ propellers and checkingof nodes, subassemblies and deck-stabbing guides ofoffshore drilling platforms.

Through direct interaction with several U.S. ship-builders and a naval architect a separate list of possibleapplications of photogrammetry was compiled.2 Thesepossible applications were judged to be not feasible orunproductive, uncertain as to productivity, or likely to be

quite productive. They are so classified and described inChapter 3.

Of the most promising applications four were actuallydemonstrated. These areChapter 2.

1.4 Summary Opinions1.4.1 Conclusions

comprehensively described in

a.

b.

c.

d.

e.

f.

g.

h.

Principal areas of application of photo-grammetry are in the- design and erectionphases of shipbuilding, although, scatteredapplications exist in other areas.

The fully analytical method of photo-grammetry is by far the most versatile forhandling a variety of measurement prob-lems. For the most part these applicationsare concerned with dimensioning large steelunits.Precise measurement of large steel units asthree-dimensional entities provides a wealthof data which can be judiciously analyzedwith an eye toward avoiding costly movesand/or lifts when corrective action may berequired.

Moreover, the fully analytical method ap-plied to such units does not require theground surveys for control nor extraordi-nary provisions for locating and framing thecamera that have been reported3 for differ-ent methods and techniques.

As a corollary, interference with ongoingShipyard work is minimal.

Equipment, software and proceduresneeded for routine application of fullyanalytical photogrammetry are state of theart and can be employed immediately by anyshipyard on a subcontract basis or throughimplementation of an in-house capability.

Analog photogrammetry is the second mostuseful method to shipbuilders.

Analog photogrammetry can best be applied in design functions which require graphicaloutput such as arrangement drawings oras-built drawings.

2

i. Although hardware for analog photogram-metry is state of the art, procedures to befollowed for different measurement prob-lems require development and refinement.

Semi-analytical photogrammetry is the leastapplicable. However, it may be very usefulfor developing digital representations ofmachinery space models for use in auto-mated design systems.

k. Accuracies of measurements derived from asound photogrammetric scheme are entirelycompatible with shipyard needs.

1. The cost and turnaround time for photo-grammetric surveys cannot generally becompared to present procedures since thephotogrammetric method can accommodaterelatively difficult situations and provide alarge amount of data: i.e. photogrammetryshould generally be considered as a supple-ment to, rather than a replacement for,current shipyard measurement procedures.

1.4.2 Recommendations:

a. Shipyards should be made cognizant of theimmediate capacity for fully analyticalphotogrammetry to accurately survey largesteel units. This should not only be donethrough distribution of this report, but alsothrough short personal presentations withinshipbuilders’ facilities.

b. At the same time, possibilities for other ap-plications of photogrammetry foundthrough the literature search and developedby this study should also be presented.

c. Should a shipbuilder elect to use photo-grammetry for a series of alike ships. seriousconsideration should be given to a combinedin-house/subcontract arrangement whereinthe shipbuilder takes the photographs and aphotogrammetrist performs the photo-grammetric work.

d. [f a shipyard elects to develop an in-housecapability for photogrammetry. technicalimplementation should be directed by aphotogrammetric consultant who shouldalso be retained for on-call service for aperiod after start-up.

f. However, more detailed study of the poten-tial for analog photogrammetry as an aid todesign of machinery spaces and development of as-built drawings is clearly war-ranted.

e. Further investigation into applications offully analytical or semi-analytical photo-grammetry is not necessary.

3

FIGURE 2-1: Blowout View of Midbody Showing Locations ofTargets. The side shell targets were offset from the butts by 2 inches.They were used primarily to analyze squareness of the side shells. A fewadditional targets (not shown) were placed on the side shells and maindeck to facilitate correlation of overlapping photographs. Taped dis-tances between other targets (not shown) on main deck provided scalereferences for the photogrammetric solution.

5

photogrammetric solution. Thus, Dennison Pres-a-plylabels were adhered directly to the deck to be certain thatno movement would occur.

2.1.2 Photographic ProceduresA two-man photogrammetric team arrived on site in

the early afternoon preceding the day scheduled forphotography. During the next morning, when theshipbuilder’s people were attaching targets, the photo-grammetric team set up their very basic developingequipment for on site processing. Test exposures weretaken and developed to verify camera settings.

Taking of final photographs commenced in the after-noon. The camera stations were occupied over a totalperiod of about four hours. Had the cranes on this newlybuilt shipway been complete and available, this time couldhave been reduced considerably. As it was, the overheadpictures were taken from a personnel carrier slung from aslow moving “cat” crane. Locations of the camera sta-tions are illustrated in Figure 2-2.1

FIGURE 2-2: Locations From Which Photographs Were Taken ofMidbody.

FIGURE 2-3: Typical Photograph of Midbody From Ground.

All exposures were secured with a Wild P31 cameraz

and glass plates coated with a panchromatic emulsion.The camera was simply moved from one station to thenext as the work progressed. Figures 2-3 and 2-4 areenlargements made from two of the glass plates actuallyused in the photogrammetric work.

2.1.3 Laboratory MeasurementsUpon receipt of the plates at the photogrammetrist’s

facility, they were examined with an ordinary magnifyingglass to locate target images. Each image was circled on the “glass” side of each plate. A few of these locationswere also numbered according to a previously devisednumbering scheme which assured that each target wouldalways have the same number regardless of which plate

upon which it appeared. Each plate was then measured ona Kern MK2 comparator. 1

2.1.4 Data ProcessingMeasurements made on the photographic plates were

processed through a set of computer programs whichtriangulate the location of each target. The particularmethod used encompasses all photographic measure-ments in one grand simultaneous or “least squares” solu-tion so as to obtain an overall best fit of all optical raysintersecting their respective targets.2

The scale of the photogrammetric solution wasroughly 1:1 relative to the actual scene photographed. Atthe conclusion of the intersection process the relativethree dimensional locations of the targets were estab-lished. Then, these data were scaled to an exact 1:1 rela-tionship using distances taped between targets on the maindeck for the scale reference. Finally, as a mathematicalby-product of the solution, the tolerance of each of thethree determined coordinates for each target was ex-tracted.

With the digital model in hand a variety of analyses arepossible and many of these were exercised to develop theas-built configuration of the midbody. Specific computa-tions carried out were:

a. Distances between selected pairs of targets werecalculated to provide information regarding halfbreadths, deck heights and the squareness of theside shells and transverse sections.

b. A best-fit plane was passed through targets lo-cated at the tank top in the transverse sections todetermine whether there was any warp in the tanktop. See Figure 2-5.

FIGURE 2-6: Deviation of Targets Representing Midbody VerticalCenterline. Arrows depict directions to target centers from the best-fitplane. Numbers are distances in 16ths of an inch.

c .

d.

e.

f.

Targets representing the ship’s centerline wereanalyzed in terms of their locations (athwartship)relative to a plane which would be perpendicularto the plane determined in “b” above. Thisanalysis would reveal systematic skewing fromthe ideal perpendicular condition as well as warpof the longitudinal vertical section. See Figure2-6.A best-fit plane was passed through targets in thetransverse section of the forward end to deter-mine how closely to a true plane the section hadbeen trimmed. See Figure 2-7.A best-fit plane was also passed through targets inthe transverse section of the after end but with thecondition imposed that this plane must also beparallel to the plane determined in “d” above.This permits determination of whether the trans-verse sections are skewed relative to each other.See Figure 2-8.Half breadths and deck heights (in the form ofoffsets and elevations) were-compared to designdata.

(1)(2)

(3)

(4)

(5)

(6)diagonal.

The program also provided deck heights and half-breadthswhich we routinely measure ourselves. ”

(see paragraph 2. 1.7.). Consequently, the next two dem-onstration tasks were purposely designed to incorporatevalid independent checks on the photogrammetric results.

2.1.6 Time and Cost AnalysisCosts for the photogrammetric work as well as for

shipyard support were closely monitored. In Table 2.1efforts of the photogrammetrist are presented in terms ofman hours and dollars so as to reflect a production situa-tion wherein a shipyard contracts for a total photogram-metric service. The shipyard’s effort is in Table 2-2.

FIGURE 2-7 Deviation of Forward Transverse Section of Midbodyfrom Best-Fit Plane. .’A’. and ● ’F” respectively denote whether the buttis aft or forwad of the best-fit plane. Numbers are distances in 16ths of aninch.

F1GURE2-8: Deviation of After Transverse Section of Midbody fromthe Plane Parallel to the Plane Best Fit to the Forward TransverseSection. “A” and “ F" respectively denote whether the butt is aft orforward of the best fit plane. Numbers are distances in 16ths of an inch.

8

The breakdowns given in Tables 2-1 and 2-2 show thatthe shipyard was responsible for fabrication and attach-ment of the targets. This could have been accomplished bythe photogrammetrist.1 Most importantly the selection oftarget locations was based upon mutual decisions.

2.1.7 Suggestions Relative to Future WorkOriginal planning of this demonstration task was done

under the assumption that work on the midbody would becompleted before the photographic effort commenced.Thus it was felt that the magnets would be well suited forattaching and removing the targets. However, some workwas in progress while the pictures were being taken and a

FIGURE 2-9: Examples of Best-Fit Circles to Turn of Bilge. In theupper diagram the best-fit circle was allowed to seek its own radiuswhereas in the lower diagram the design radius was enforced. Arrowsdepict directions to steel surface (built to the molded surface) from thebest-fit circle. Numbers are distances in 16ths of an inch.

number of the targets were disturbed. The magnets werealso less than ideal for positioning on beveled edges. Thefollowing is recommended:

a.

b.

If shipyard work is in progress, use self-adhesivetargets directly on flat surfaces. Irregular or bev-eled surfaces should be covered with a square of16-gauge sheet metal tack-welded in place.If shipyard work is not in progress, magnets maybe used even for some irregular surfaces providedthe targets are held squarely. Flat-faced clips,Figure 2-10, can be used to square off a bevelededge.

TABLE 2-1: Photogrammetrist’s Man Hours andCost for Dimensioning the Midbody2

1. Labor Man Burdened “‘Hours Ra te3 C o s t

a .

b.

c.

d.

e.

f .

Project planning and 44 19 $ 836coordinationincluding visit toshipyard (1 man)Prepare and ship 35 12 420equipment and roundtrip travel (2 men)On-site preparation 46 12 552and photography(2 men)Prepare photographic 12 12 144 plates (1 man)Measure plates (sub- — — 562contracted)Data processing and 61 19 1.159reporting ( 1 man)

Total Labor . . . . . . . . . . . . . . . . . ..$3.673

2. Expenses

a .

b.c.

d.

Transportation and per $ 934diem (initial visit andon-site work)Materials (photographic plates, etc.) 114Computer 6 3 0

Miscellaneous (air freight, etc.) 128

TABLE 2-2: Shipbuilder’s Effort in Support ofDimensioning the Midbody

1. Labor

a.

b.

c.

d.

e.

Task Man Hours

Fabricate targets 6

Attach targets 12

Target maintenance 8

Assistance in personnel 8carrier for overheadphotographs

Crane operator 4

TOTAL . . . . . . . . . ..38 manhours

2. Miscellaneousa. Target materials $ 2

b. Crane 4 hrs.

FIG URE2-10: Flat-faced Ctip For Squaring Off Beveled Edges.

2.2 Measlurement of a Hold Receiving a Spherical LNGTank3

This demonstration was conducted at the QuincyShipbuilding Division of the General Dynamics Corpora-tion where a series of eight LNG tankers were beingconstructed. The method of containment utilized theMoss-Rosenberg spherical tank design. For this particu-lar demonstration the shipyard primarily desired to checkthe circularity of the 120-foot diameter cylindrical steelsupport skirt which is to receive the first LNG tank. Thisstructure must mate with a cylindrical aluminum supportskirt which is attached to the equatorial ring of the spheri-cal tank. As will be seen. additional data were also de-veloped from the same photographs.

2.2.1 PreparationA visit to the shipyard was conducted to determine

dimensioning requirements and to observe the hold so asto become familiar with the physical conditions underwhich the work would be performed. Because of the needfor precise dimensions in all directions and to relate pointson the weather deck opening some 30 feet above the top ofthe cylinded to the cylinder itself, the fully analyticalmethod of photogrammetry using convergent photo-graphs was chosen for executing the task. Subsequent tothis initial visit to the shipyard, planning of photo-grammetric field procedures commenced. This planningeffort was as described for the midbody in paragraph2.1.l.

An obvious photographic plan would have been totake pictures from above the weather deck looking downinto the hold so as to see the entire top end of the cylinder.To do this, however, would have required the use of theshipyard’s 1,200-ton crane. Since the photogrammetristdid not wish to disrupt any shipbuilding work in progress,an alternate photographic scheme which did not requirethe crane was devised. Although this alternative involveda greater photogrammetric effort, the trade-off was favor-able because there was no interference with ongoing ship-yard work.

In advance of the scheduled date for photography GDQuincy fabricated about 150 targets in accordance with a design specified by the photogrammetrist. Each target wasa 1 1/2-inch square cut from a 16-gauge aluminum sheet, one side of which was painted flat black. A 3/4 inch diameterwhite Dennison Pres-a-ply label was adhered to the painted side and a double sided pressure sensitive adhe-sive material was attached to the unpainted side.

The target location scheme is illustrated in Figure

l Because targets were being disturbed by ongoing work, continuous checking of the target locations was required. AlSO see paragraph 2.1.7.2 Except for the self adhesive bull’s-eyes, targets were made from scrap.3Also, see subparagraph 3.2.3c.4Hereinafter the word “cytinder” is substituted for “cylindrical steel support skirt. ”

10

2-11. GD Quincy attached 96 targets to the top of thecylinder on its inner surface, 28 on the sloping bulkheadsat the bottom of the hold, 8 on the longitudinal and trans-verse bulkheads and 22 on the edges of the weather deckopening. Targets placed on the top of the cylinder wereintended primarily to represent its shape. But, they werealso aligned with webs of vertical stiffeners, on thecylinder’s outer surface, so that arc distances between thestiffeners could be determined.

Targets on the sloping bulkheads were to be used intwo ways. Distances were to be taped between four pairsof targets for the purpose of providing scale referencesnecessary for the photogrammetric solution. Taped dis-tances between other selected pairs of targets were to beused as a check on the accuracy of the photogrammetricwork. As it was wise to have redundant references, dis-tances were also taped between the four pairs of targets onthe vertical bulkheads.

FIGURE 2-11: Locations of Targets and Camera Stations in an LNGSpherical Tank Hold. Targets aligned with the cylinder’s inside top edgeare also aligned with the webs of the outside vertical stiffeners. Thelongitudinal and transverse bulkheads, and the eight targets located onthem, are not shown. See Figure 2-15 for locations of targets on the edgeof the weather deck opening.

Targets on edges of the weather deck opening incombination with those on top of the cylinder were fordetermining whether their respective planes were skewed.In addition, the weather deck targets were to be refer-enced to the axis of the as-built cylinder to determine thedistances between the axis of the cylinder and the edges ofthe weather deck opening.

All but the weather deck targets were attached onlyby means of the adhesive on the backs of the targets.Flat-faced clips, Figure 2-10, were used on the edges ofthe weather deck opening. Targets were then adhered tothe faces of the clips such that target centers were alignedwith the bottom surface of the weather deck, i.e., theas-built surface which corresponded with the designmolded surface. .-

2.2.2 Photographic ProceduresBasic developing equipment was set upon site in the

dark room used for the shipyard’s X-ray exposures. Testexposures were immediately processed to verify appro-priate camera settings. Final photographs were taken bothfrom the weather deck and from the bottom of the hold.Eight photographs horn the weather deck were “looking”down and across the hold whereas four photographs fromthe bottom of the hold viewed across and out of the hold.Figure 2-11 illustrates the arrangement of the camera sta-tions. A total of 60 minutes was required to secure all 12photographs once the targets were in place.1 All photo-graphs were exposed on glass plates coated with a panchromatic emulsion. Figures 2-12 and 2-13 were madefrom two of the glass plates which were actually used inthe photogrammetric solution.

2.2.3 Laboratory Measurements and Data Process-ing

Preparation and measurement of the photographicplates proceeded exactly as described for the midbody inparagraph 2.1.3. Photogrammetric computations were asdescribed in paragraph 2.1.4.

Once the photogrammetric computations had con-structed the digital model of the cylinder and weather deckoverhead, as represented by the targets, the followingcalculations were performed:

a. A best-fit plane was passed through the top of thecylinder to determine how closely it conformed toa true plane. Actually, the top had not yet beentrimmed but this computation was nonethelessperformed to demonstrate the technique and alsoto establish a plane which could be related to theweather deck above. See Figure 2-14.

b. A best-fit plane was passed through targets on theweather deck but with the condition that it beparallel to the best-fit plane through the top end ofthe cylinder. This calculation was designed toreveal skewing between the two planes. See Fig-ure 2-15.

c. A best-fit circle was passed through targets on thetop of the cylinder. This computation was de-signed to establish the’ ‘out of roundness” of thecylinder as well as its best-fit inside radius. SeeFigure 2-16.

d. As a by-product of step “c” above the center ofthe best-fit circle was determined and projectedupward, perpendicular to the plane best fit to thetop of the cylinder. Its intersection with the paral-

lel plane best fit to the weather deck allowsaccurate prediction of clearances between thespherical tank, during its installation, and theweather deck opening.

PORT

AFT

FIGURE 2-14: Deviation of Top of Cylinder From the Best-Fit Plane.Values are in 16ths of an inch. Plus means steel is above the best-fitplane. Some systematic “highs and lows’” are evident: excess stock hadnot been trimmed before pictures were taken.

FIGURE 2-12: Typical Photograph From Weather Deck of Hold for aSpherical LNG Tank.

FIGURE 2-13: Typical Photograph From Tank Top of Hold for aSpherical LNG Tank.

skewing relative to the top of the cylinder is evident. Large gaps indistribution of data resulted from targets being obscured by staging.

e. Finally, arc distances between targets on top ofthe inner surface of the cylinder were computed.These were used to calculate distances betweenthe vertical stiffeners on the outside of thecylinder. 1

2.2.4 Evaluation of ResultsCertain distances obtained from the photogrammetric

calculations were forwarded to GD Quincy. The shipyardthen taped five of the same distances for comparison;2 allwere about 45 feet in length. The average difference be-tween the two sets of distances was 1/16 inch and themaximum difference was 3/16 inch. This agreement is con-sidered to be exceptionally good because:

b. Taped distances were determined by summationof segments with the total lengths recorded to thenearest 1/16 inch. Ordinary shipyard procedurewas used in taping the distances.

c.

d.

The distances were taped about 10 days after thephotographs were taken and the surveyors tapingthe distances had no knowledge of the photo-grammetric results.The differences between photogrammetric andtaped distances encompass errors in both mea-surement processes.

The shipyard’s analysis of the entire photo-grammetric product generated a letter which stated”. . .very complete information in a concise manner with excel-lent accuracy. We are most pleased with the photogram-metric technique. . ."

2.2.5 Time and Cost AnalysisVery close accounting was kept of- man hours and

costs involved in dimensioning the hold. Man hours andcosts for the photogrammetrist’s effort, as presented inTable 2-3, have been adjusted to reflect a production situa-tion wherein a shipyard subcontracts for the service. Theshipyard’s effort in terms of man hours is given in Table2-4.

Of the four photogrammetric demonstrations con-ducted for this project, this demonstration was the onlyone which was executed as if it were indeed a productionsituation. Preparation of the photographic plates, measur-ing of the plates, and data processing were completed tothe point that initial results could be telephoned to theshipyard starting on the eighth calendar day after takingthe photographs. A complete letter report including tabu-lations of data, sketches and a narrative were forwarded tosketches and a narrative were forwarded to the shipyardon the 12th calendar day. This is considered to be a typicalturnaround for a project of this magnitude when workingon a one shift per day basis. Two and even three shiftscould be employed to speed the process.

2.2.6 Suggestions Relative to Future WorkAs indicated in Figure 2-15 a number of targets on the

edge of the weather deck opening could not be triangu-lated because their images failed to appear on two or morephotographs. This was the result of obstructions to viewcaused by staging. For more complete data, staging infront of the targets could have been temporarily removed.An alternate possibility would be to offset the targets, inelevation, by known amounts.2.3 Measurement of a Transverse Bow-Section

At GD Quincy an LNG tanker hull was made avail-able to demonstrate photogrammetric dimensioning of atransverse section having high degrees of curvature. Thesection measured was about 135 feet in breadth by 82 feetin height and was to receive a 900-ton bow erection unit.

TABLE 2-3: Photogrammetrist’s Man Hours and Costs forDimensioning a Hold to Receive a Spherical LNG Tankl

1.

a.

b.

c.

d.

e.

f .

g.

2.

a .

b.

c.

d.

Labor

Man BurdenedHours Ra te2 C o s t

Initial visit to ship- 24 19 $ 456yard (1 man)

Project planning 13 19 247(1 man)

Prepare and ship equip- 36 12 432ment and round triptravel (2 men)

On-site preparation and 32 12 384photography (2 men)

Prepare photographic 8 12 96plates (1 man)

Measure plates — — 351(subcontracted)

Data processing and 48 19 912reporting (1 man)

Total Labor . . . . . . . . . . . . . . . . . . . . . . ..$2.878

Expenses

Transportation and per $1,074diem (initial visitand on-site work)

Materials (photographic 104plates, etc.)

Computer- 730

Miscellaneous (air freight, etc.) 103

TOTAL EXPENSES . . . . . . . . . . . . . ..$2.011$4.889

PRoFIT@ 15% . . . . . . . . . . . . . . . . . . . . . . 733

TOTAL . . . . . . . . . . . . . . . . . . . . . . . .. $5.622

1. Labor

a .

b.

c.

d.

Task

Fabricate targets3

(including flat-faced clips)

Paint targets

Carpenters - to extendstaging and attach targets

Surveyors -

Man Hours

12

2

7

(1) accompany. photo-grammetric team4

(2) measure distances5

11

24

TOTAL . . . . . . . . . ..56man hours

2. Miscellaneous

I a. Target materials . . . . . . . . . . . . . . . . . . . ..$10(est.) -

1Where travel is involved. a Florida based photogrammetric firm working at CD Quincy is assumed.2Rates vary among firms.3This is partially a non-recurring cost since the flat-faced clips may be reused.4An observer was required for security reasons due to use of a camera within a shipyard where Navy work was in progress.5This does not include check distances since these would not normally be taken in a production situation.

14

which pictures could be taken. In fact, the only logicalvantage points (save for the use of a crane or speciallyerected staging) were across the top of the dock gate.From there it was more convenient to expose a series ofhighly overlapping nominally oriented photographs in-stead of a regularly arranged group of stereo pairs.

In advance of the date scheduled for photography GDQuincy fabricated about 75 targets similar to those usedin the LNG tank hold (see Section 2.2.1) except that thediameter of the bull’s-eye was reduced to 1/2 inch. Figure2-17 illustrates the targetting scheme wherein about onehalf the targets defined the shell plating. Flat-faced clips,Figure 2-10, were fixed to the plating and each target wasthen adhered to the face of a clip such that the center ofthe target was aligned as near as possible with the“molded” edge of the shell. The remaining targets wereadhered to the transverse bulkhead in a well distributedpattern in order to provide additional common points

FIGURE 2-18: Typical Photograph of the Transverse Bow Section.

among the overlapping photographs. A number of thesetargets were also aligned with the bulkhead centerlineas a reference for subsequent calculation of halfbreadths. One target was located at the intersection of thebulkhead centerline and the l-foot waterline to provide areference for computing elevations.

Shipyard surveyors taped several distances betweenpairs of targets. Some were provided to the photogram-metrist for use as scale references in the photogrammetriccalculations while the remainder were withheld to checkthe photogrammetric results.

2.3.2 Photographic ProceduresPhotographs were taken from the top of the dock gate

as shown in Figure 2-17. The total elapsed time for ob-taining all 9 exposures was 59 minutes. All were exposedon glass plates with the same camera1 used for the mid-body and cylinder surveys. Similarly, on-site processingwas performed to assure satisfactory exposures. Figure2-18, made from one of the photographs, illustrates thedegree of curvature of the shell plating.

2.3.3 Laboratory Measurements and Data Pro-cessing

Preparation and measurement of the photographicplates proceeded exactly as described for the midbody in

1A Wild P31 camera as illustrated in Appendix B, Figure B-2.

paragraph 2.1.3. Photogrammetric computations were asdescribed in paragraph 2.1.4. Because of the high overlapof the pictures some individual targets appeared on asmany as seven frames and, therefore, were triangulated byas many optical rays.

Upon completion of the photogrammetric computa-tions the digital model of the transverse section was trans-lated and rotated into the ship’s coordinate system accord-ing to the following conditions:

a.

b.

c.to have zero values in the fore-aft direction, againin a best-fit sense. 1

2.3.4 Evaluation of ResultsPhotogrammetrically derived coordinates of the

targeted points defined the shape of the shell. To comparethis as-built shape to the design, it was first necessary forthe shipyard to obtain offsets and elevations for a fictitiousreference frame since the master butt was located betweenframes. Then the differences between “as built” and’ ‘de-sign” were obtained as follows:

a.

b.

c.

d.

Offsets determined by photogrammetry were cor-rected because the actual steel at each target wasactually forward or aft of the adopted designreference frame (relative fore and aft positions ofthe targets were known from the photo-grammetric solution).

Portions of the reference frame thought to mostclosely correspond to the actual steel measuredwere plotted at a very large scale (about 3:1 rela-tive to the ship) in the vicinity of each target.

Photogrammetrically determined locations of thetargets (but corrected as described in “a” above)were superimposed upon the plots.

The perpendicular distance between the as-builtlocation of each target and the design molded hullwas simply scaled with a ruler. Figure 2-19 de-picts the overall differences found. (Their mag-

nitudes were confirmed by the shipfitters whenmating the bow unit.)

As a partial check on the accuracy of the photogram-metric solution distances between selected pairs of targetswere computed and provided to the shipyard. These werecompared to corresponding distances taped and recorded

PORT

SCALE FOR DEPARTURES

FIGURE 2-19: Deviation of Shell Plating of Transverse Bow Section.The design transverse section, corresponding to location of the masterbutt. is the reference illustrated. Each vector is perpendicular to thereference and depicts the direction to the steel surface built to the moldedsurface.

by only the shipyard on the day during which the photo-graphs were taken.

Taped distances were measured in segments andsummed to the nearest 1/16 inch to obtain total distances.Thus, combined photogrammetric and taping errors con-tribute to the differences reported.

2.3.5 Time and Cost AnalysisAn accounting of the photogrammetrist’s man hours

and costs for this demonstration is presented in Table 2-5. As in previous analyses the figures have been adjustedslightly to reflect a production situation wherein a ship---—. .yard subcontracts for a photogrammetric survey. The

. 1This was done even though the section had not been trimmed. Any small resultant error in being perpendicular to the ship’s centerline would haveinsignificant effect on computed offsets and no effect on computed elevations.2The average difference of 2/16 inch is reasonable but the maximum difference of 5/16 inch seems too large to be attributable to ordinary measuring errors.It is probably invalid because this difference applied to a distance which was actually a segment of a longer distance which checked within 2/16 inch.

16

TABLE 2-5: Photogrammetrist’s Man Hours and Costsfor Dimensioning the Transverse Bow-Section 1

1.

a .

b.

c.

d.

e.

f.

g.

Labor

Man BurdenedHours Rate2

— —Initial visit to ship- 24 19yard (1 man)

Project planning 16 19(1 man)

Prepare and ship equip- 36 12ment and round triptravel (2 men)

On-site preparation and 22 12photography (2 men)

Prepare photographic 4 12plates ( 1 man)

Measure plates — —(subcontracted)

Data processing and 41 19reporting ( I man)

cost

$456

304

432

264

48

189

779

Expenses

Transportation and per $1.074diem (initial visitand on-site work)

Materials (photographic 62plates, etc.)

Computer 271

Miscellaneous (air freight, etc.) 103

Total Expenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .$1.510

$3.982Profit @ 15% . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..$ 597

Dennison circular white Pres-a-ply self-sticking labelupon which a small black dot was rubbed on from a trans-

2.4.2 Photographic ProceduresA series of photographs was taken with a single cam-

era from the locations illustrated in Figure 2-20. Afterturning the model around, a similar series of photographswere exposed to aid mapping of detail from its other side.Although a rather large number of photographs were ex-posed, only six of these were ultimately used. The pur-pose in taking more photographs than absolutely neces-sary was simply to supplement, with slightly differentviews, those few photographs which would be used for.most of the mapping.

All exposures were taken on glass plates coated with a

FIGURE 2-20: Locations From Which Photographs Were Taken ofMachinery Space Model. Photographs were also taken of the outboardside. The shaded portion of model was ignored for this demonstration.

panchromatic emulsion using a Wild P31 camera. 1 Nospecial illumination was used. Ambient light through thewindows plus overhead fluorescent lights provided ade-quate, but not absolutely ideal illumination. All photo-graphic plates were developed on site using very basicdeveloping equipment and procedures. Figure 2-21 wasmade from a typical photograph. To expediteidentification of the color coded pipes as rendered in blackand white, color snapshots were also taken for referenceby the stereo plotter operator.2

2.4.3 Mapping ProceduresOne pair of photographs, corresponding to a view

from inboard looking outboard, was placed in a stereoplotter for viewing and mapping of the optical mode13

according to the following:

a.

b.

c.

d.

e.

f.

The optical model was rotated until all targets onthe wall were at the same depth as measured in thestereo plotter.

Straight-line details, such as edges of decks andbulkheads, were traced. A mechanical linkagefrom the stereo plotter to the plotting table causeda pencil to draw at the correct scale. The lineswere smoothed and inked by hand.

Next, edges and detail of fixed machinery weretraced. Again, the penciled lines were smoothedand inked by hand.

Because of the complexity of the piping only onepipe was traced at a time, with the operator refer-ring to the color snapshots as necessary. As piping drawings typically show centerlines of pipes,the actual procedure was to pencil an “eyeball”best-fit straight line through each traced edge of a pipe and then take the average of these two lines

as representing the location of the centerline. 1

The centerline was then inked and the pencilededges erased so as to avoid confusion in plottingadditional pipes.

Additional pairs of photographs were insertedinto the stereo plotter as necessary to complete alldetail. Steps “a” through “c” were repeated foreach new pair of photographs. Steps “d" through“f” were repeated for new detail only.

The mapping operation was performed by two per-sons. One operated the stereo plotter while the othersimultaneously “heavyed up” detail being traced. That is,the second person served as a draftsman for the inter-polation and drafting functions involved in steps “d”through “f” in the foregoing. Upon completion of themapping work the original drawing was aligned on a dig-itizing table for measurement of distances between thecenterlines of pipes and the deckheads or forward bulk-head.

Next, a check print was made and compared to themachinery space model. Annotations were made of miss-ing detail. extraneous detail and names of machinery anddistributive systems. Then, the photographic plates andthe original drawing were reintroduced in the stereo plot-ter and required missing detail was recovered andmapped.

Figure 2-22 is a reproduction of the final drawing. Atthe suggestion of the shipbuilder, no drafting refinementshave been added since they would provide no additionaluseful information.

2.4.4 Evaluation of R esultsThe drawing was checked both for content and for

accuracy of dimensioned distances. Detail content wasfound to be very good. Some small diameter piping, suchas used for the steam station against the forward bulkhead,could not be retrieved from the available photographs. Inboth instances it is likely that these details could have beenmapped (at least to a great extent) had it been possible toobtain photographs from above the machinery spacemodel. Nonetheless, it is anticipated that a small amountof manual finishing will always be required to complete apiping drawing prepared by photogrammetry.

Spot checks were made on the accuracy of the photo-grammetrically prepared drawing by comparing 20 of itsdimensioned distances to like distances on the shipbuild-er’s conventionally prepared drawings. Large differ-

2.4.5 Time and Cost AnalysisThis particular demonstration is difficult to analyze in

terms of time and other costs which might be experiencedby a shipbuilder in a production situation. Several factorscontribute to this uncertainty.

a.

b.

c.

d.

The task was quite unlike any other undertakenby most photogrammetrists. Some inefficiencyresulted from developing procedures as the workprogressed.

A plan view could not be obtained and only thesteam systems within a portion of the model weremapped in the elevation view.

The mapping work may have been more efficienthad it been possible to prepare a plan view first.With only photographs from the sides of themodel it was difficult to account for “hidden”pipes in common horizontal planes.

Certain fixed costs apply regardless of how muchof a model is mapped.

1 lt should be mentioned that tracing edges of a pipe does not strictly produce its exact diameter to scale. This is because a photograph is a perspectiveview and optical rays through the lens tangent to a curved surface intercept a chord. Also. this interpolation process would not be necessary with a wireand disc representation of piping.

2See Appendix C. Part 1, Article B-2.

19

20

e. Finally, the portion of the model chosen fordemonstrating the photogrammetric method isconsidered to be the most complex. Therefore itprobably represents the most difficult and expen-sive area to map.

Costs for preparing the elevation drawing shown inFigure 2-22 were adjusted as best as possible to reflect aproduction situation and are contained in Table 2-6. Usingthese adjusted costs as a basis, costs for a plan view wereestimated and are also included.

2.4.6 Suggestions Rlative to Future WorkExecution of this demonstration was undertaken with

full knowledge that neither the available machinery spacemodel nor the contemplated photogrammetric processescould be classified as being ideal. Moreover, it was alsorecognized that preparation of just a drawing of steamsystems was merely touching upon the full potential forphotogrammetric measurements as an adjunct to designmodeling. 1 The intent in proceeding with this demonstra-tion was to develop a basic example of the usefulness ofphotogrammetry for the design of piping systems. If ashipyard were to consider implementation of design mod-eling combined with photogrammetric dimensioning, thebasic required improvements are:

a. The model should be constructed with a viewtoward accessibility for photography. Preferablythe model would be more modularized so that itcould be disassembled at least in units bounded bytransverse bulkheads. Also, separations at thecenterline plane and the ability to separate upperand lower deck levels would be very helpful.

b. The use of clear plexiglas for the sides, bulkheadsand decks should be minimized to the greatestpossible extent. Its light reflecting and refractioncharacteristics are disruptive.

c. For greater measurement accuracy a wire anddisc representation of piping systems should beused. This would also allow mapping to progressmuch more rapidly since a pipe could be mappedsimply by identifying a few points along its path.

d. More uniform lighting should be furnished. espe-cially inside the model to minimize glare andshadows.

e. Materials with dull finishes and flat coatingsshould be substituted for polished materials andglossy paints respectively. to further minimizeglare.

Total Expenses . . . . . . . . . . . . . . . . . . . . . . . 1,250

$5,458Profit @ 15% . . . . . . . . . . . . . . . . . . . . . . . . . $ 819

TOTAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..$6,277 5

1 It was observed in June 1976 that Odense Steel Shipyard Ltd., in Denmark, successfully combined pipe-systems design and modeling to create 1:15scale models of ship machinery spaces without preparing Piping systems or composite arrangement drawings. With input from production people,sufficient model detail is provided to identify multi-ton piping units, consisting of a number of different systems. which are planned for shop assembly.Dimensions are manually lifted from such models and recorded on a simple isometric sketch for each planned piping subassembly (on the average a subassembly consists of about eight pipe pieces and eight fittings). Each isometric. sketched on art 81/2X 11 inch sheet, contains all material requirements,pipe piece details, and ship location references. Thus, each isometric combines all design data required for production, a significant amount of completedplanning, and all information necessary for the remaining planning and scheduling functions. Further, each isometric sketch. being sufficient, is thencoded for computer preparation of material lists, pipe bending instructions, work orders, schedules, etc. More details of this Odense design/modelingprocess are to be incorporated in the report for the study “Advanced Pipe Technology,” also part of the National Shipbuilding Research Program, whichis expected to be published in eariy 1977.2For steam systems in the portion of the machinery space model shown in Figure 2-22.3Where travel is involved, a Florida based photogrammetric firm working at Sun SB & DD is assumed.4Rates vary among firms.5Does not incIude shipyard’s efforts for project coordination and proof reading of drawing check copies.

22

f. The camera should not be located in any horizon-tal or vertical plane containing groups of pipes.Avoidance of these “critical planes" wouldminimize hidden pipes on the photographs.

g. Consideration should be given to a larger photo-graphic scale to increase the photogrammetric ac-curacy. Necessarily, this must include considera-tion of interrelated factors such as focal length,relative aperture, ability to change focal length forshort focus and depth of field.

DEMONSTRATION

Measurement of CompleteMidbody Section

Measurement of Hold toReceive Spherical LNGTank

Measurement of aTransverse Section

Measurement ofMachinerySpace Model

Another improvement may be the use of color film.This is still unproven but ordinary color snapshots arenonetheless desirable for reference.

2.4.7 Long Range PotentialAs mentioned above, if the wire and disc modeling

technique is used, a pipe can be mapped by identifying afew points along its path. Since a stereo plotter measuresin all three dimensions simultaneously and since each axiscan be digitized (a very common practice), the pointsdefining a pipe can be digitized.

TABLE 2-7: Summary of Demonstrated Applications

SECTION DATA DEVELOPED ACCURACY TESTS COST

2.1 Half breadths and deck N/A1 -$6,301

heights with comparisonsto design; squareness ofsides and transversesections; planarity oftank top; deviation ofcenterline plane; devi-ations of trimmed sectionsfrom true planes; skew oftransverse sections; devi-ations of bilge turns fromtrue circles

2.2 Deviation of top edge ofcyl- 5 taped distancesinder from true plane; ave. differencedeviation of weather deck 1/16 inch; max.overhead from true plane difference 3/16

parallel to top edge of cyl- inchinder; out-of-roundness of cyl-inder; distances from axisof cylinder to points onweather deck opening; arcdistances between verticalstiffeners on outside ofcylinder

2.3 Shape of shell platingand deviations fromdesign

2.4 Elevat ion drawingof steams systems

in portion of model

5 taped distancesave. difference2/16 inch; max.difference 5/16

inch

20 design distances -ave. error 11/8inch; max. error25/8 inch at scaleof ship3

$ 5 , 6 2 2

$4,579

N/A2

1Validity of tests uncertain: see paragraph 2.1.5.2Only partial drawings were compiled: see paragraph 2.4.5.3lncludes errors in model as well as photogrammetric error. Accuracy can be improved.

23

Such digital representations could be manipulated toautomatically plot system arrangement drawings, com-posites or isometrics at any desired scale. Also, pipe bend-ing details could be automatically generated as has beendemonstrated elsewhere.l Ultimately, the digital datacould be merged with other automated design systems.For these potential applications it is clear that photo-grammetry could serve as an excellent input “device”which would permit a combined pipe-systems designer/model maker to put his inherently interference-free pipingarrangements into a computer.

2.5 Summary of Demonstrated ApplicationsThe principal finding from the four shipbuilding ap-

plications of photogrammetry that were demonstrated aresummarized in Table 2-7, page 23.

1See Appendix C. Part 1, Article B.3.

24

3. A SURVEY OF PHOTOGRAMMETRY INSHIPBUILDING

To determine past and current applications as of theinitial stages of this project (February-May 1975) athorough review of pertinent literature was conducted inparallel with a worldwide mail solicitation of individualsthought to have knowledge of such applications. In addi-tion, interviews with U.S. shipbuilders generated furthersuggestions for other potential applications.

3.1 Literature Search and Direct Mail SolicitationCoincident with an extensive literature search in-

quiries were sent to 68 persons in 59 countries. 1 Bothefforts were sufficiently broad in scope to include indus-tries having industrial functions similar to those in ship-building. The results indicate that photogrammetry hasnot been widely adopted.

The few instances wherein photogrammetry isroutinely used for shipbuilding include a Norwegianpatented service for producing sounding tables for LNGspheres, an East German process to aid the manufactureof ships propellers and measurement of drilling platformcomponents in Britain. Aside from shipbuilding, photo-grammetry has apparently been adopted by one companyas the preferred means for “lifting" dimensional data fromscale models of chemical plants. Some automobile man-ufacturers use photogrammetry to develop contours andcross sections of design models which would be roughlyequivalent to developing a lines drawing from a model of aship’s hull.2 However, producing a shell expansion whichis a related application is not potentially productive be-cause of the efficient computerized method already in use.

While they have not widely adopted photogrammetry,shipbuilders have shown considerable interest in its po-tential. This has been manifested by several papers, inves-tigations and experiments which were uncovered duringthe literature search and mail campaign. Pertinent ab-stracts are in Appendix C. A brief summary categorizedby shipbuilding functions is contained in Table 3-1.

3.2 Review of Suggestions from U.S. ShipbuildersDuring the first several months of this project the

photogrammetrist was required to acquire a basic under-standing of shipbuilding functions. This was accom-plished by background reading, and to a greater extentthrough visits to several U.S. shipbuilding yards and to anaval architect’s office. While conducting these visitsideas were solicited for dimensioning tasks which might besuited to photogrammetric techniques of measurement.All phases of shipbuilding were considered. This section

1Twenty-seven responses were received (40%).

TABLE 3-1: Applications of Photogrammetry Revealedby Literature Search and Direct Mail Solicitation

ShipbuildingFunctionDesign

Fabrication

Erection

OutfittingShip Repair

Miscellaneous

Application

1. Measuring bow and stem waveforms under simulated speed anddepth of water conditions in a towingbasin.2. Determining trajectory and speedof a test model under varying poweroutput and rudder position.3. Dimensioning from scale models.4. Measuring cavitation on modelpropellers.5. Measuring cavitation on propellerat sea. `

6. Development of pipe bending data.7. Dimensioned drawings of vendors’equipment.1. Checking propellers during andafter their manufacture.2. Checking shape of rolled andformed shell plates.3. Checking dimensions of nodes fordrilling platforms.1. Dimensioning of three-dimensionalerection units (ships and drilling plat-forms).2. Predicting fit of large erection units.3. Determining as-built shape of hulls.4. Preparing sounding tables for LNGspheres.

None1. Dimensioned drawings of a histori-cal ship.1. Determining trajectory and speedof a ship during launching.2. Measuring subsidence of ways dur-ing launch.3. Determining trajectory and speedof a ship upon impact with a pier usedto pivot the ship for alignment with itsberth.4. Dimensioning of a ship’s propellerafter a period of service.

2Photographs of a hull model used to predict hydrodynamic performance characteristics could be cheaply made and easily filed. If necessary at a futuredate photogrammetric processing would produce the lines of the model tested for comparison IO those used for construction of a ship.

25

contains the substance of such discussions. It also illus-trates the degree of shipbuilding know-how acquired bythe photogrammetrist.

3.2.1 Applications in Design

a. Drawings from Design ModelsSome people believe that basic design,

especially of distributive systems, should befirst accomplished with a design model. 1

This is in contrast to the more conventionalapproach wherein the design is developed inthe form of arrangement drawings first andthen, perhaps, tested for interferences bymeans of a model. If the “design modeling”approach were to be adopted, the problemarises as to how the design, as conceived inthe model, can be accurately captured andconveyed to all persons requiring knowl-edge of the design. In posing this question itis assumed that the model cannot be the onlymeans for communication, and obviouslysome drawings based upon the model arerequired. Photogrammetry is seen as an ac-curate and efficient means for producingsuch drawings.2 A description of the pro-cess, demonstrated with a ship’s machineryspace model, is contained in Section 2.4.

b. Shell Plating ExpansionIt has been customary in the past to first

prepare a shell plating model, upon which allbutts and seams are scribed. To determinethe developed form of each plate by photo-grammetry would be a rather straightfor-ward process by the semi-analytical method.Stereo pairs of the shell plating model wouldbe introduced into a stereo plotter and theedges of each strake would be digitized in X,Y, Z stereo plotter coordinates. Subsequentcomputer processing would project the datafor each plate to a plane to realize the de-veloped shape. The same procedure couldconceivably be followed for surfaces of veryhigh curvature such as propeller bossings.

The photogrammetric solution seems tobe naturally suited to this task, but the appli-cation has not been given serious considera-tion since modem computerized design sys-

C.

terns have all but eliminated the practice ofpreparing shell plating models.

Shape of Shell Plating Around HawsepipeIt is usually necessary to build an

anchor handling model to assure a properanchor handling arrangement. Once a finaldesign has been established, it must be con-verted into offsets and elevations for fabri-cation and erection purposes. One proce-dure presently employed uses a moldingtechnique. A silicone rubber3 casting ismade directly from the lower hawsepipe andshell bolster area of the model. This castingis then used as a pattern to reproduce theoriginal model in a plastic substance4 thatcan be readily sawn. Offsets are measuredfrom the parallel sections obtained bysawing.

Analog photogrammetry is very wellsuited to this dimensioning task. A stereopair of photographs of the anchor handlingmodel would be placed in a stereo plotterand the surface of the resulting opticalmodel would be traced along any desiredwaterline or frame. Output from the stereoplotter would be a graphical presentation ofthe shape of each line relative to the ship’scenterline. Actual offsets at intersections offrames with waterlines would be shown innumerical form.

But, while photogrammetry can easilyaccomplish such measurements it is notlikely to be more productive than existingprocedures. Moreover, the frequency of oc-currence of this dimensioning requirementis rather low. Even weighed against the factthat photogrammetry can produce moredata, this potential application of photo-grammetry seems more academic than pro-ductive as seen from the shipbuilders’ pointof view. However, consideration should begiven to just photographing the shell bolsterarea of the constructed ship. Then, if ananchor handling problem occurred, photo-grammetry could be used to analyze the as-built surface.

1MachineW space design modeling. as performed by Odense Steel Shipyard Ltd.. is described in footnote 2 on page 22. Also. this opinion is expressed inChapter II, Section F. page 6. "Use of Scale Models as a Management Tool." National Shipbuilding Research Program report by the U.S. Departmentof Commerce Maritime Administration in cooperation with Todd Shipyards Corporation, May 1974.2Also. see Appendix C. Part 1. Articles B. 1 and B.2.3"Silastic.”4A mixture of polyester auto body putty and polyester resin.

26

3.2.2 Applications in Fabrication

a.

b.

Measurement of Flat Panels and Sub-Assemblies

As flat panels and sub-assemblies arefabricated, it is common practice, as a mat-ter of quality control, to check their dimen-sions. For the most part this checking isaccomplished simply by measuring the lengths of sides and diagonals. This ratherbasic checking procedure is viewed as beingadequate in most instances. Since relativelyfew dimensions are required, usually in ashort period of time, and since the neededdimensions are readily obtained with a steeltape, photogrammetry cannot compete.

Measurement of Small Erection UnitsSeveral shipbuilders have expressed

some concern with annoying fitting dis-crepancies which often occur. The difficultyseems more predominate when curved sur-faces are involved. It has been suggested bymore than one shipbuilder that photogram-metry be applied to this problem as a qualitycontrol measure.

There is no question that fullyanalytical photogrammetry can be used fordimensioning individual modules and per-haps even to the extent of predicting the fitbetween modules. However, the questionarises as to the value of such detailed infor-mation unless there is time to plan remedialwork such as compensation in erection unitsnot already built. Save for a gross discrep-ancy, it is likely that such information wouldonly serve to predict the need for extraman hours during erection.

It is accepted that the described dis-crepancies do exist, but it seems more pru-dent to direct shipbuilders’ efforts towardbetter dimensional controls in the burningand fabrication stages. If photogrammetry isto be applied at all, it should only be in atemporary capacity for checking three-di-mensional modules as an aid to perfectingother quality controls.

3.2.3 Applications in Erection

a. Measurement of a Complete MidbodyIn the “jumboizing” process entire

midbody sections are inserted into existing

1A description of this process, as demonstrated for shipbuilders is contained in Section 2.1.

27

ships. Because of the large size of such unitsit is difficult by conventional means to de-termine, in detail, the as-built configuration.Necessary measurements include:(1)

(2)

(3)

(4)

(3

(6)

(7)

half breadths and heights which locatethe intersections of decks, shell andother structural members which con-tribute to longitudinal strength,

shape such as turns of-the-bilge,

comparisons of items (1) and (2) to de-sign data,

squareness of all sides,

deviations of transverse sections fromideal planes,

racking of the entire unit,

parallelism of the forward and aftertransverse sections.

As a practical matter items (6) and (7)are of major interest since racking and paral-lelism cannot be readily ascertained by con-ventional means.

Although emphasis is placed upon a de-termination of racking, this is not tominimize the capability of photogrammetryfor generating data for each of the otherlisted items from the same set of photo-graphs. It is also important to recognize thatcomprehensive dimensioning of a midbodyunit1 is practically identical to requirementsfor dimensioning mating faces of separatelybuilt ship halves.

In the context of jumboizing it has beensuggested that photogrammetry might beused to measure the receiving ship after ithas been cut. These measurements wouldbe compared to data previously developedfor the midbody so as to predict its fit. Thisconcept is precisely that described in sub-paragraph 3.2.3.b which pertains to mea-surement of mating faces of a ship built inhalves. However, insofar as “jumboizing”is concerned, the concept is not practicalwhen the joining process commences withina few hours of cutting the ship. There is notsufficient time to permit measurement byphotogrammetric methods. However, con-sideration should be given to just photo-graphing the exposed transverse sections

after the ship is cut. Then if there are unan-ticipated matching problems caused by dif-ferences between the ship as built and theowner furnished “as-built” drawings, factsregarding the differences could be estab-lished by photogrammetric processing.

b. Measurement of Mating Sections of a ShipBuilt in Halves

Measurement of mating sections wouldpermit predicting their fit well in advance ofmoving either or both sections. A totalmeasurement program should include virtu-ally the same data described in subpara-graph 3.2.3.a for measuring a ship’s mid-body.

There is no question photogrammetricmeasurement of mating sections of shiphalves and predicting their fit would be pro-ductive. If such predictions saved even oneworkday in a large capacity building dockthe costs for applying photogrammetrywould be more than offset.

One shipbuilder has suggested a varia-tion in the application of photogrammetry tomeasuring mating sections of ship halves.This variation would have photogrammetryapplied to at least one half of the ship afterthe trim line has been established by a laser,but before trimming has been performed.Thus a check would be made on the perpen-dicularity of the trim line relative to theship’s centerline and base planes. Thereason for such a check is not lack ofconfidence in the laser; it is the physicaldifficulty of establishing true references. Itwas also suggested that if one ship half hasbeen built neat, the possibility exists for ac-curately plotting its as-built dimensions andfor adjusting the trim on the other ship halfto suit.

c. Measurement of a Hold Receiving an LNGTank l

One design for transporting LNGmakes use of a spherical tank which, whenlowered into a ship, rests upon a cylindricalsupport skirt. Fabrication and erection ofthis cylinder is quite criticaI considering that

28

d.

the diameter is about 120 feet and that itmust match a portion of the cylinder at-tached to the sphere’s equatorial ring. Acomprehensive measurement programshould provide the following data:

(1)

(2)

(3)

(4)

(5)

(6)

radius of the circle which best fits a rep-resentative number of measured pointsaround the top of the cylinder,

departures of the measured pointsaround the top of the cylinder from thebest-fit circle,

distances between vertical stiffeners at-tached to the outside of the cylinder,

departures of measured points from aplane best fit to the top of the cylinder,

skew of the plane of the weather deckrelative to the plane best fit to the top ofthe cylinder, and

upward projection of the best-fit circlefor the purpose of predicting clearancesbetween edges of the weather deckopening and the spherical LNG tank.during its installation.

Consideration has also been given tomeasuring the mating surface of that portionof the support skirt attached to the equator-ial ring of the sphere. The data sought wouldconsist of the planarity of the mating surfaceand its circularity; see items (1) and (2)above. With both parts of the skirt so mea-sured, their fit could be predicted before thesphere was installed.

Measurement of a Drilling Rig

One shipbuilder has reported annoyingfit discrepancies when floating deckhouses onto supporting columns of semi-submersible drill rigs. At present steel-tapemeasurements are taken across the verticalcolumns which are each about 18 feet indiameter, not perfect cylinders, and about

90 feet apart. Aside from access problems,there is doubt whether the tapes are uni-formly tensioned for reliable measurements.

The shipbuilder suggested that photo-grammetry be used to determine relativeheights and orientations of the columns.This information would be of value fortrimming the columns to a common planeand for predicting their fit to the deckhouse.This potential application is viewed as beingfeasible and productive. The approachwould make use of the fully analyticalmethod.

3.2.4 Applications in Outfitting

a. Measurement of Propeller and RudderKeyways

To provide totally useful data for key-ways it is necessary to dimension both themale and female components. While it isfeasible to measure the male component byphotogrammetry, the female presents se-vere space restrictions which cannot behandled by ordinary photographic tech-niques suitable for photogrammetric dimen-sioning. Moreover, large keyways are be-coming obsolete.

b. Measurement of Surface Contact of Reduc-tion Gears

Present practice for determining thecontact of reduction gears makes use of a“blueing in” process. The face of each toothon one gear is coated with a substance whichupon rotation of the gears is deposited on themating teeth. The amount of contact be-tween the mating teeth along the pressureline is indicated by the presence of depositedcoloration.

The actual amount (percent) of contactmade is apparently determined via humanjudgment. Photographic procedures aresometimes used, but again, the determina-tion of the amount of contact is still partiall ysubjective.

Analysis of the photographs could cer-tainly be made quantitative by applyingsimple photogrammetric principles. How-ever, if there is no need for greater accuracy

relative to present methods the added effortfor a photogrammetric analysis is not war-ranted.

c. Measurement for Shaft Alignment

To set foundations for propulsion-shaftspring bearings, it is necessary to determinethe height relationship between the bottomof the shaft alley and the reference line. Inthe interest of productivity two shipbuildershave suggested that it would be desirable toweld spring-bearing foundations in place be-fore the erection units in which they arelocated are welded together. Some shipbuilders already do this by simply allowinggreater tolerances and planning for the useof thicker chocks during installation of eachbearing.

In concept photogrammetry would beused to survey the bottom of the shaft alleyand to dimension the individual erectionunits. Subsequent digital processing would“connect” the units together so that the en-tire shaft alley could be analyzed as a singleunit. It is doubtful, however, whether theapplication of photogrammetry is practicaldue to space restrictions on taking photo-graphs in a shaft alley. Even if it were possi- ble to acquire suitable photographs, the digi-tal connection of erection units could verywell be unlike that which is actuallyachieved.

d. Dimensioning for Walkways in MachinerySpaces

Often when only one ship of a type isbuilt, or when a ship is the first of a series,the final layout of walkways in machineryspaces evolves from design drawingsmodified many times as the machinery outfitprogresses and interferences are disclosed.It has been suggested that, followinginstallation of the main machinery compo-nents, photogrammetry might accelerate theentire process.

The photogrammetric approach is notparticularly straightforward since photo-graphs would generally be taken from un-usual points of view. Numerous obstruc-tions of view so typical of machinery spacesalso complicate the photogrammetric solu-tion. The manual effort involved in taking

29

and analyzing the photographs may not becompetitive with present procedures. How-ever, more detailed study should be under-taken before firm conclusions are reached.

e. Measurement for Locating Framework forLNG Insulation

To position the insulation system in ahold designed for a Conch LNG tank aframework of balsa/plywood support panelsmust first be laid out along the junctionswhere the sloping and vertical bulkheadsintersect one another and the tank top. Theframework is also needed along the tops ofthe vertical bulkheads. Once in place, thisframework ideally conforms to the designshape of the inner hull irrespective of theas-built configuration.

An original suggestion regarding thepossible use of photogrammetry was as ameans for checking the locations and“heights” 1 of the support panels after instal-lation. Subsequently, this concept was re-vised to employ photogrammetry prior toplacing the panels rather than after the fact.Specifically, photogrammetry would beused to accurately dimension the as-builtinner hull of the ship. Further data process-ing would then be performed to generate theneeded elevation data for each main groundand also the location of its main stud. Thephysical location of the stud on the actualsteel would be found by scribing two orthree arcs from nearby photogrammetrictargets using radii determined from the dataprocessing.

Two potential photogrammetric solu-tions for this task were developed. One solu-tion employed a large number of targets onthe inner hull and the fully analytical methodof photogrammetry whereas the second so-lution used fewer targets and a combinationof fully analytical and semi-analytical photo-grammetry.

The estimated cost and turnaround foreither solution were found to be about thesame as that experienced by the shipyardusing strictly manual procedures. Thus, in

the instance considered, a photogrammetricsolution was not productive. However, theshipyard did feel that more serious consid-eration would have been given to a photo-grammetric approach had it been presentedbefore the yard had undertaken and mas-tered the manual technique.

f. Development of Sounding Tables for LNGTanks

Fully analytical photogrammetry can beemployed to precisely determine the as-builtconfiguration of a tank as represented statis-tically by a few hundred targets attached tothe shell. In the case of a spherical tankphotographs would be taken from the insideof the tank.2 Photographs of a tank of theConch design would likely be taken fromoutside due to numerous internal obstruc-tions to view. In either case, the digitalmodel of the tank would be integrated inincrements required for the sounding tableswith appropriate corrections applied forservice temperature and internals.

3.2.5 Applications in Sea Trials

a. Measurement of Hogging and Sagging

Photogrammetric measurement of thehogging and sagging of a ship at sea couldconceivably be accomplished by semi-analytical or fully analytical methods. Ineither case at least a few targets would beattached to the hull of the ship. Photographswould be taken with two synchronizedcameras located in a chase ship or aircraft.Photogrammetrically determined locationsof the targets using any pair of photographswould represent the shape of the ship at theinstant of exposure. Comparison of resultsfrom different pairs of photographs wouldindicate the changing deformation of theship.

For normal situations it is not believedthat there is a real need for such data. How-ever, this is another case where photographscould be made and filed in anticipation ofsubsequent events which may justify photo-grammetric processing.

1 Heights in this context means distances between main grounds of the panels and the actual steel. Distances vary due to deviations of the as-built steelfrom the design.2In January 1976 General Dynamics Corporation, Quincy Shipbuilding Division awarded a contract to develop sounding tables for 40 LNG tanks of theMoss-Rosenberg spherical design. Fully analytical photogrammetry was the specified method of measurement.

30

f. Measurement of Bow Wave and Wake

To monitor bow wave and wake thephotogrammetric approach would be to ex-pose photographs with two synchronizedcameras from above the ship. Analogphotogrammetry would be utilized to mapthe wave forms. Again, this is a case wherephotographs could be maintained on file inthe event that a future circumstance justifiesprocessing.

3.2.6 Applications in Ship Repair

a. Measurement for Repair of Hull Damage

Repair of hull damage is oftentimes de-layed when as-built or design drawings ofthe ship are not available. One method forobtaining such data is to hang a probing de-vice over the undamaged side in order tomeasure the required offsets. Sketches mayalso be prepared from measurements madeinside the ship.

Analog or semi-analytical photo-grammetry is very well suited to gatheringdata from outside of the hull. Moreover, thephotogrammetric approach could be moreproductive and would be more accurate thanpresent methods. To relate the photogram-metrically determined shape to the ship’sbase and centerline planes, a minimum ofthree points whose offsets and elevationsare known (or can be readily determined)must be visible on the photographs.

3.2.7 Miscellaneous Applications

a. Development of As-Built Drawings 1

Virtually any area which can be readilyphotographed is conducive to the prepara-tion of as-built drawings.2 These may berelatively simple areas such as panels con-taining a multiplicity of gauges, switches,indicator lights, etc., or more complexspaces containing piping, ducting and wire-ways. Also, dimensioned drawings of ven-dors’ equipment can be similarly prepared

prior to delivery of the equipment to theshipbuilder.

In new construction of larger shipscomposite drawings are theoretically updated to as-built status upon completion ofthe ship. In practice this is oftentimes notsufficiently done. Whether photogrammetrycan play a useful role in bringing the draw-ings up to as-built status is questionablefrom cost effectiveness and productivityconsiderations; a manual ship check may bejust as efficient.

When only one ship, or the first of a newdesign is built, oftentimes much of the piping, ducting, gauges, switches, etc. are fieldrun or placed. Thus, each such ship canserve as a full scale design model for prepar-ing drawings of such details. Photogram-metry may be very useful for this purpose.

In ship upgrading work the first courseof action is to conduct a ship check to de-termine existing as-built conditions. Often-times as-built drawings are not available andit becomes necessary to develop them anew.Typically, this involves a team of men whotake ordinary snapshots and preparesketches in the ship for use in preparingsufficiently detailed arrangement drawings.In the most difficult upgrading work severalman years may be expended just to obtainthe as-built drawings needed for planning. Ifas-built drawings are required after the up-grade work is completed, an additional andeven greater effort is oftentimes necessaryto produce the final drawings.

The greatest application for photo-grammetrically prepared as-built drawingsappears to be in mapping of pipes, ducts andwireways in overhead spaces. For thephotogrammetric method to be productive itis felt that it is necessary to develop a pro-cedure built around a simplified stereo plot-ter which can be operated with minimumtraining by technician level personnel withina shipyard or naval architect’s office. A tol-erance at the scale of the ship on the order of

1 Although normall y considered a design function. preparation of as-built drawings has been separated from the section treating potential applications ofphotogrammetry in design. It is felt that photogrammetric methods may be suited to dimensioning vendors equipments and to ship upgrading work as wellas to the more conventional requirement for as-built drawings of new ships.2Machinery spaces are excepted since it is not generally possible to ‘stand off’ with a camera to obtain unobstructed views.

31

3.3 Productivity of Applications Suggested by U.S. Ship-builders

The following categorizations are in the context of theapplication descriptions contained in Section 3.2 preced-ing:

LIKELY TO BE PRODUCTIVE• Ship’s midbody for “jumboization.”

• Mating sections of ships built in halves.

● Ship’s hold for a spherical LNG tank.

● Structure of semi-submersible drilling rigs.

• Sounding tables for LNG tanks.

● Repair of hull damage.

● Piping arrangement drawings from design mod-els.

• As-built drawings.

PRODUCTIVITY UNCERTAIN● Grounds for LNG tank insulation systems.

ž Machinery space walkways.

• Hogging and sagging.

● Bow wave and wake factors.

NOT LIKELY TO BE PRODUCTIVE● Shell-plate expansion.

● Shell plating around hawsepipe.

● Smaller erection units.

● Flat panels and subassemblies.

• Propeller and rudder keyways.

● Surface contact of reduction-gear teeth.

• Foundations for propulsion-shti spring bearings.

4. OPTIONS FOR APPLYINGPHOTOGRAMMETRY IN A SHIPYARD

To adopt photogrammetry as a measurement tool ashipbuilder has first to select from three alternatives: de-velop a total in-house capability, retain a photogramme-trist by subcontract, or use a combined in-house/sub-contract arrangement. Each are objectively discussedherein. All estimates are based upon the monetary valuesexisting in March 1976.

4.1 Total in-House CapabilityA totally self-sufficient photogrammetric unit re-

quires capital for hardware, software and training ofselected personnel. Estimated costs, itemized in Table4-1, depend upon whether a shipyard desires a capabilityfor analog, semi-analytical, fully analytical photogram-metry or a combination of these.

As implied in Table 4-1, rather extensive trainingwould be required for implementing an in-house capabil-ity. Initially, as a minimum, a two-man team dedicated tophotogrammetric surveying would be required. Prefera-bly, one person should have prior experience in photo-grammetry while the other should have experience inshipbuilding. If semi-analytical or fully analytical photo-grammetry is desired, computer support on a timely basiswould also be necessary. A start-up period of about threemonths should be allowed before production work com-mences.

4.2 Total Subcontract ArrangementTotal subcontracting of photogrammetric work is bas-

ically the method by which demonstration surveys de-

scribed in Chapter 2 were performed. Although the shipyards did participate to the extent of fabricating targets,attaching targets and taping distances, these functionswould normally be performed by the photogrammetricfirm. However, there are situations (e.g. scheduling andsafety restrictions) wherein attachment of targets and taping of distances would be best performed by the shipyard.

Typical costs involved for subcontracted photo-grammetric surveys of some shipbuilding applications arecontained in Tables 2-1 through 2-6. Appendix D listsfirms which provide photogrammetric services.

4.3 Combined In-House/Subcontract Arrangement

If there is a series of alike ships to be built the shipyardmay wish to consider assuming responsibility for just thefield effort. Primarily this would involve taking the photo-graphs and taping control distances but may also includetargeting if fully analytical photogrammetry is employed.Photographs secured by the shipbuilder would then beforwarded to a photogrammetrist for analysis. The firstone or two surveys, however, should be entirely overseenby the photogrammetrist in order to instruct shipyardpersonnel in procedures to be followed for subsequentsurveys. These procedures would be quite straightfor-ward and would not vary from survey to survey.

To iIustrate the combined concept, consider the costsfor photogrammetrically surveying a hold in an LNG ship;see Tables 2-3 and 2-4. If it is desired to survey 40 holds1

and if the shipyard fabricated and installed targets andtaped distances, the photogrammetrist’s charge would beabout $5,600 for the first survey and about $4,400 for each

TABLE 4-1: Estimated Start-Up Cost for Implementing Photogrammetry In-House2

circa March 1976

COST ITEM METHOD OF PHOTOGRAMMETRY

1.

2.

3.

4.

5.

ANALOG SEMI-ANALYTICAL FULLYANALYTICAL

Camera(s) $4.000- $ 20,000 $4,000-$20,000 $13,000-$20,000

Stereo plotter 33.000- 105.OOO 46,000- 118,000 N/A

Comparator N/A 24,000- 35,000

Computer programs N/A 2,000- 30,000 40,000

Manuals and in-plant 14,000 14,000 12,000training

TOTALS . . . . . . . . . . . . $51,000-$139.000 $66.000-$182,000 $89,000-$107,000

1 Eight ships with 5 holds each.2 In preparing these estimates. only hardware and software deemed suitable for shipbuilding have been considered so as to present realistic estimates.Ranges in hardware costs reflect differences in sophistication and capability of instruments. See Appendix D for suppliers of photogrammetric hardwareand services.

33

of the 39 subsequent surveys.1 Thus, the total cost wouldbe approximately $177,200. If the shipyard were to pur-chase a camera and undertake all field work for, say, 38surveys (the first two would be supervised by the photo-grammetrist, the photogrammetrist’s charge would be

survey and $2,500 for the remaining 38. In this case then,the total subcontract cost would be roughly $105,000. Tothis add $16,000 for the camera and photographic supplies.The result is a reduction in subcontract costs of about$56,200.

The combined approach may also be attractive to ashipbuilder since it allows total flexibility for taking pic-tures. The need to coordinate timing with thephotogrammetrist’s field team would be eliminated.

34

APPENDIX A

GLOSSARY OF PHOTOGRAMMETRIC TERMS

Analog Photogrammetry-the method in which a stereopair of photographs is used in a stereo plotter to createa three-dimensional optical model and measurementsof the optical model are produced as a direct graphicalpresentation.

Comparator- a device for measuring relative locations ofimages of points on a photograph; the instrument con-sists of an x-axis, a y-axis, a stage upon which thephotograph is lain, a measuring reticle and viewingoptics.

Contour- a line of constant value within the scene; forexample, the shape of the hull at any waterline.

Convergent Photography -the procedure in which the optical axes of the photographs are purposely inclinedrelative to one another and with respect to the scene;normally done to increase the accuracy of fullyanalytical photogrammetry.

Cross Section - a trace of shape of the scene, usually in adirection perpendicular to a reference line; for exam-ple, the shape of the hull at any frame.

Digital Model -a scaled three-dimensional rendition ofthe scene photographed generated by digital pro-cessing of comparator measurements of images ofspecific points within the scene or by point by pointdigitizing of an optical model in a stereo plotter; thedigital model consists only of points in the scenewhose images are measured.

Distortion -see lens distortion.

Fully Analytical Photogrammetry -the method in whichimages of specific points within the scene are meas-ured directly on single photographs, stereo pairs orconvergent photographs using a comparator; a three-dimensional digital model may be formed by computa-tion and then further processed by digital means forfinal presentation of graphical and/or numerical re-sults.

Intersect -see intersection.

Intersection -in fully analytical photogrammetry, theprocess of digitally projecting optical rays from cor-responding images of the same point on two or morephotographs to their intersection at that point withinscene; the process is equivalent to triangulating thelocation of a point through measurements of horizon -tal and vertical angles with a transit or theodolite fromtwo or more known locations.

Lens Distortion -geometric errors introduced into the re-corded photographic image by the inability of the lensto allow a point in the scene, the optical center of thelens and the image of the point to lie on a straight line.

Measuring Reticle -a dot or cross hair within a stereoplotter or comparator used to sight upon specificpoints of interest or to trace continuous details such ascross sections, profiles and contours.

Model -see optical model.

Optical Axis -that line which passes through the opticalcenter of the lens and is perpendicular to the focalplane of the camera.

Optical Model -a scaled three dimensional rendition ofthe scene photographed, created in a stereo plotter byprojecting light through the original negatives ortransparent copies thereof; the separately projectedimages are viewed with special optics so as to fuse likeimages and create a perception of depth.

Optical Ray -a straight line emanating from a photo-graphic image passing through the camera lens andcontinuing to the corresponding point in the scene.

Photogrammetric Measurement -a measurement madedirectly on a photograph or in a three-dimensionaloptical model.

Photogrammetry -the science of extracting meaningfultwo or three-dimensional measurements of a scenefrom one or more photographs of the scene.

Profile -a trace of shape of the scene, usually in a direc-tion parallel with a reference line; for example, theshape of the hull at any buttock.

Reticle -see measuring reticle.

Scene -the portion of the real world as imaged by a cam-era; in shipbuilding the scene is usually a specificobject such as a ship’s hull, a midbody section or ascale model of a machinery space.

Semi-AnalyticaI Photogrammetry -the method in which astereo pair of photographs is used in a stereo plotter tocreate a three-dimensional optical model and photo-grammetric measurements of the optical model areprocessed digitally to obtain final results.

Single-Image Photography -the procedure in which suc-cessive photographs overlap one another by only asmall amount; any portion of the scene is imaged(essentially) only on a single photograph.

software -computer programs.

A-1

Standard Deviation -a statistical measure of error com-monly used in photogrammetry; as a practical matter,one standard deviation multiplied by 2 would beroughly equivalent to tolerance.

Stereo Pair -two photographs whose optical axes arenominally parallel and which are exposed such that(about) 50% or more of the scene on one photographalso appears on the other photograph.

Stereo Plotter -a projection device used to create athree-dimensional optical model from a stereo pair ofphotographs; see optical model.

Tolerance -maximum allowable or expected plus-minuserror of a measurement. For universal applicability inphotogrammetry, the value may be expressed in partsper thousand of the major dimension of the scenephotographed.

Triangulate -see intersection.

APPENDIX B

A LAYMAN’S EXPLANATION OFPHOTOGRAMMETRY

1 DEFINITION OF PHOTOGRAMMETRY

Photogrammetry is the process of extracting mean-ingful two or three-dimensional measurements of a scenefrom one or more photographs of the scene. The derivedinformation may be in the form of coordinates of points,distances, outlines of features or shapes of surfaces. Aglossary of photogrammetric terms is presented in Ap-pendix A, to which the reader is urged to refer whilereading further.

2 GENERAL CONCEPTS AND METHODS OFPHOTOGRAMMETRY

The desired dimensional data of the scene photo-graphed can be obtained through several combinations ofphotogrammetric processes; the various possibilities areillustrated in Figure B-1. In most instances the combina-tion of processes used for a given practical situation isdictated by the nature of the object, the required accuracyof results and the type of data presentation desired.

2.1 Camera(s)A photogrammetric camera is of robust construction

to ensure a constant relationship between the lens and thefocal plane. The lens is virtually free of distortion. Film orglass plates are used to register the exposure; somecameras can accept both glass plates and film while otherscan accommodate only one or the other.

FIGURE B-1: Typical Combinations of Photogrammetric Processes.

WILD P31 ZEISS JENA UMK

WILD C40 CARL ZEISS SMKFIGURE B-2: Examples of Single and Dual Cameras.

When the scene to be photographed is stationary, asingle camera may be used to take photographs from oneor more camera locations (i.e., the camera is moved be-tween exposures). Alternately, two or more cameras maybe used and if the scene to be photographed is in motion,the camera shutters must be synchronized. Oftentimes apair of cameras are rigidly mounted at a known separationupon a bar with the optical axes of the cameras parallel.Such an arrangement can be used for static and dynamicsituations and also introduces some simplification in thephotogrammetric process owing to the known relationshipbetween the cameras. Typical cameras are shown in Fig-ure B-2.

2.2 Classification of Photographic CoverageThe size and shape of the scene to be photographed

and the accuracy desired of the final results generallydictate the type of photographic coverage employed.

2.2.1 Single Image PhotographyWhen the area of interest lies in a plane, a single

photograph or a series of slightly overlapping photo-graphs may be used. See Figure B-3.

B-1

FIGURE B-3: Illustrating Single Image Photography. Photographsoverlap only slightly. Any portion of the scene is seen (essentially) only on one photograph. A single camera may be used to take all photographs.

FIGURE B-4: Illustrating Stereo Photography. Adjacent photographsusually overlap by at least 50 percent and camera axes are nominallyparallel. When viewed in a stereo plotter the imagery common to adja-cent photographs is perceived in three dimensions. A single camera maybe used to take all photographs.

2.2.2 Stereo PhotographyWhen the scene is three-dimensional and contour,

cross section or profile data are desired, two or morephotographs are taken such that their optical axes areparallel and such that each photograph overlaps itsadjacent photograph by 50% or more. Any two adja-cent photographs are called a stereo pair. See FigureB-4.

2.2.3 Convergent PhotographyWhen the scene is three-dimensional and the ac-

curacy requirements are stringent, two or morephotographs may be taken with their optical axes pur-posely inclined with respect to one another and to thescene photographed. Such photographs are calledconvergent photographs. See Figure B-5.

FIGURE B-5: Illustrating Convergent Photography. Camera axes arepurposely inclined to one another and with respect to the scene. A singlecamera may be used to take all photographs.

2.3 Classification of Photogrammetric Measurement “‘Photogrammetric measurements 1may be made di-

rectly on the original negatives (or reproductions there-from) or in a three-dimensional model created from astereo pair.

2.3.1 Direct Measurements of PhotographsWhen photogrammetric measurements are to be

taken directly on a photograph, details of the scene forwhich data are desired must be well defined on thephotograph. To provide the needed definition con-trasting markings (e.g., paint) may be used or targetsmay be placed at specific points of interest within the scene. The photographs, or more precisely, relativelocations of thedetails of interest on a photograph, aremeasured on a comparator. A comparator for mea-suring single photographs simply consists of a fixed

1 Photogrammetric measurements are not the same as the desired measurements of the scene photographed. The desired measurements of the scene arederive from the photogrammetric measurements.

B-2

measuring reticle, a fixed optical viewing system, astage upon which the photograph is lain and whichmay be moved along two axes. Typically, the move-ment of the stage is digitized with a least count of 1micrometer (0.00004 inch). Typical comparators formeasuring single photographs are shown in FigureB-6.2.3.2 Measurements in a Three-Dimensional Model

A stereo pair of photographs may be used tocreate a three-dimensional optical model of the scene(or portion thereof) which is common to the twophotographs. The two photographs can be likened tothe images received by the human eyes. Since thesame scene is viewed from two slightly different per-spectives by the eyes, there is a sense of depth perception (i.e., three-dimensionality). The same effect canbe achieved with a stereo pair of photographs. Origi-nal negatives or transparent reproductions thereofmay be placed into side-by-side projectors and when

KERN MK 2

KERN PG3FIGURE B-7: Examples of Stereo Plotters for Measuring OpticalM o d e l s .

2.4

the projected images are viewed through special op-tics, the observer actually sees a three-dimensionaloptical reproduction of the scene photographed. Thisoptical reproduction is commonly referred to as amodel since it is a true rendition of the scene photo-graphed. However, the model is usually at somethingless than one-to-one in proportion to the scene photo-graphed; hence, the term “model.” The instrument inwhich the model is viewed and measured is called astereo plotter. See Figure B-7.Classification ofData Processing and Presentation ofResultsData processing and presentation of results derived

from photogrammetric measurements are dictated by thetype of photogrammetric measurements taken.

2.4.1 Direct Graphical PresentationWhen an optical model is cross sectioned,

profiled, or contoured, or when outlines of features aretraced, these traces of form may be produced directly.This is accomplished on a drafting table which islinked to the measuring reticle of the stereo plotter.The stereo plotter operator moves the reticle aboutthe surface of the optical model and the pencil au-tomatically follows and draws the movements. Thisgraphical presentation may be mechanically or elec-tronically scaled up or down as desired. That is, the

B-3

graphical plot need not be the same scale as the opticalmodel. This combination of measurement and presen-tation is known as analog photogrammetry (see FigureB-1).

2.4.2 Digital Calculations Followed by Numericalandlor Graphical Presentation

When locations of individual points are measuredwithin an optical model (as opposed to cross section-ing, profiling or contouring) some digital processingusually follows. As examples, this may involve trans-formation of photogrammetric coordinates measuredin the model to a ship coordinate system or the mea-sured points may be subjected to a surface fittingroutine. In the former case the results would mostlikely be presented in numerical form; for instance,offsets and elevations, half breadths, deck heights,etc. In the latter case the results would probably bedisplayed graphically; for instance, a contour mapdepicting departures from an ideal surface. In eithercase, the combination of measuring an optical modeland presenting results after some intermediate digitalcalculation is known as semi-analytical photogram-metry (see Figure B-1).

When measurements are taken directly fromphotographs, data processing is always required. As aminimum, a digital model must first be produced. Thisdigital model is equivalent to the optical model withthe exception that the digital model consists only ofthose specific points whose images were measured onthe photographs. Point data of the digital model areusually subjected to further processing, as just de-scribed, prior to final presentation of results. Thecombination of taking photogrammetric measure-ments directly on the photographs and subjectingthese to digital processing is known as fully analyticalphotogrammetry (see Figure B-1).

3 CONTROL FOR PHOTOGRAMMETRY

Since photogrammetry is an indirect measurementprocess, it is necessary to have a known reference uponwhich the photogrammetric solution may be based.

3.1 Scale ReferencePhotogrammetric measurements are necessarily at

the scale of the photographs or of the optical or digitalmodels formed from the photographs. The scale relation-ship between the photogrammetric measurements and thescene photographed must be precisely established. Thismay be accomplished within the scene or at the camerastations.

3.1.1 Scale Reference Within the SceneScale reference within the scene is readily created

by placing a survey rod within the field of view of the

camera(s), by stretching a steel tape across or next tothe scene or by precisely measuring the distance be-tween two well defined points marked within thescene.

3.1.2 Scale Reference at the Camera StationsThe scale of a photogrammetric solution may also

be determined if the camera stations are separated byknown distances. As mentioned earlier, some dualcamera systems are purposely manufactured with thisconcept incorporated. Distances between camera sta-tions may also be measured with a steel tape.

3.2 Absolute ReferenceBefore describing methods for establishing an abso-

lute reference it must be mentioned that some measure-ment tasks do not require an absolute reference.Specifically, if distances between points (for example, halfbreadths and heights) are the desired end product, thenthere is no need to relate the photogrammetric coordinatesystem to any other absolute coordinate system (such as aship coordinate system). Also, if it is desired, for example,to determine the shape of a curve or surface, here again, anabsolute reference is not required. An absolute referenceis required only when the results must be reported in aspecific fixed coordinate system. It is also noteworthy thatscale may be implicit in an absolute reference and, there-fore, a scale reference need not always be separatelyestablished.

3.2.1 Absolute Reference Within the SceneAbsolute reference within a scene is usually es-

tablished by discrete points within the scene whosecoordinates in the reference coordinate system areknown. For example, to relate photogrammetricmeasurements of a section of ship’s hull to the shipcoordinate system, at least three discrete points ofknown offset, elevation and longitudinal distancewould permit a determination of the relationship. It isnot necessary, however, that all three coordinates ofany single reference point be known. Points whosecoordinates are only partially known can be used too.For instance, as a minimum, two points of knownoffset and longitudinal distance plus three points ofknown elevation will also suffice.

3.2.2 Absolute Reference at the Camera StationsTo establish an absolute reference at the camera

stations it is necessary to survey the locations of thecamera stations (position and elevation) and to deter-mine the directions in which the cameras are pointed,relative to the reference coordinate system.

B-4

4 ACCURACY OF MEASURING BYPHOTOGRAMMETRY

4.1 Definition of AccuracyIn photogrammetry, accuracies are usually expressed

in terms of standard deviations or parts per thousand ofthe distance from the camera(s) to the scene. These con-ventions have been discarded in favor of a scheme whichis more practical for day to day estimating and which isbased upon the criterion of tolerance (i.e., maximum ex-

applicability, it is necessary to express tolerance in termsof unitless ratios. Specifically , tolerance is expressedherein as parts per thousand of the major dimension of thescene photographed and applied to the accuracy of singlepoints.

To illustrate the method of expressing tolerance, sup-pose that it has been established (by this report) that aparticular photogrammetric method can yield tolerances

the scene photographed. Assume also that the scene is atransverse section and that the breadth of the section is150 feet while the height is 100 feet. The “major dimen-sion” is taken to be the diagonal distance which, by calcu-lation, is 180 feet. Division of the diagonal distance by

There are two significant peculiarities of accuraciesobtainable by photogrammetry. First, as already indi-cated, tolerance as defined above applies to the absoluteaccuracy of locating an individual point. For all practicalpurposes all points are located with essentially the sameaccuracy. Thus, points which are physically close to oneanother will have less relative accuracy than points whichare more widely separated. The second peculiarity is thatthe accuracy of about 50% of the points will actually bemore precise than the indicated tolerance by about a factorof 2.1

4.2 Accuracv of Analog PhotogrammetryAs analog photogrammetry deals primarily with

stereo pairs of photographs, the accuracy of results in thedirection to and away from the observer is less than inplanes parallel to the observer. This is directly a result ofthe fact that the separation between the cameras is usuallysmall relative to the distance from the cameras to anypoint in the scene, thereby forming a slender trianglewhose apex in the scene is not strongly determined by theadjacent sides. Another determining factor affecting accu-

racy is the quality of the stereo plotter employed. Theserange from relatively unsophisticated “approximate solu-tion” instruments to very sophisticated computer con-trolled stereo plotters.

For approximate reference the following table may beused as a guide for tolerances to be expected of analogphotogrammetry.

Tolerance in Planes Tolerance in DirectionParallel to Observer2 to and From Observer3

n o t u s e a b l eto to

4.3 Accuracy of Semi Analytical PhotogrammetryTolerances expected of the semi-analytical method

are about the same as the maximum figures for analogphotogrammetry. This is because the distance betweenthe cameras is usually near the maximum which can beaccommodated by the stereo plotter and also because astereo plotter of medium to high quality is normally used.

4.4 Accuracy of Fully Analytical PhotogrammetryBecause the photogrammetric measurements are

taken directly from the photographs and processed en-tirely by digital means, the fully analytical method affordsgreater opportunity for sophistication (for example, tocorrect for small errors such as lens distortion, film defor-mation, etc.). For this reason, the fully analytical methodis preferred when point data of high accuracy are needed.When stereo pairs are used in the fully analytical process,

The accuracy of fully analytical photogrammetrymay be further enhanced by the use of convergent photo-graphs and more sophisticated computer programs. Thisenhancement inaccuracy is achieved by intersecting eachpoint in the scene from many camera stations. Moreover,the triangles formed by the camera stations and points inthe scene are no longer slender and the tolerance to andaway from the observer tends toward equality with thetolerance in the planes parallel to the observer. Toler-

5 ADVANTAGES AND DISADVANTAGES OFMEASURING WITH PHOTOGRAMMETRY

5.1 Advantagesa. Complexity of shape or detail of the scene is not

restrictive.

1Throughout this report tolerance has been equated to two standard deviations. Theoretically it is more correct to equate tolerance to three standarddeviations. The former is justified as a practical matter since only 4.6 out of 100 dimensions would theoretically exceed the stated tolerance. Experiencehas shown further that the maximum error is actually about 2.5 standard deviations rather than three standard deviations. Moreover, a "typica]”accuacy is about one standard deviation.2I ncreases with more sophisticated stereo plotters.31ncreases with increasing separation between the cameras, but the distance is ultimately limited by mechanical restrictions of the stereo plotter.

B-5

b. Relative directions to all points of interest withinthe field of view of a camera are recorded instan-taneously with a single opening of the shutter.(This is in contrast to time consuming measure-ment of horizontal and vertical angles with atransit or theodolite.)

c. All data (photographs) can be obtained in a shortperiod, thereby minimizing the effect of thermaland gravity induced changes within the sceneupon the consistency of the final results.

d. The amount of data to be extracted from thephotographs does not pose severe time and costlimitations.

e. The process is virtually noncontacting.

f. Due to the short period of time required for takingthe photographs, there is minimum interferencewith ongoing work. In some instances work maynot be interrupted at all.

g. Dynamic situations can be accommodated withthe use of two or more cameras having syn-chronized shutters.

h. The orientation or attitude of the scene is of noconsequence. Photographs can be taken fromabove, below or to the sides so long as there isroom to stand off with the camera.

i. With the fully analytical method of photogram-metry, the accuracy can be varied over a widerange to suit the particular need.

j . The photographs constitute a permanent recordand may be reused to check results or gatheradditional data.

k. Photographs may be taken and archived for anypotential need to produce data at a future point intime. This applies to potential litigation as well asto pure technical requirements.

5.2 DisadvantagesMeasurement by photogrammetry has drawbacks as

well as advantages. Notable disadvantages are:

a. Results of photogrammetric measurements arenot produced instantaneously. Depending on thedifficulty of the task considered, the time requiredto produce final results may range from severalhours as a minimum to several days. An extremesituation might involve several weeks.

b. Useful photographs cannot be taken in rain, snowor fog.

1These considerations are treated in greater detail in Chapter 4.

B-6

c. Measurements in very confined spaces are gener-ally not practical due to the large number ofphotographs needed to cover the entire scene.Also, the depth of focus of the camera lens maynot be able to accommodate varying distances topoints within the scene.

d. The first cost for an in-house capability rangesfrom $50,000 to $180,000 depending on the extentof the capability desired. 1

e. Again, depending on the complexity of the capa-bility desired, a photogrammetrist or special train-ing of a shipbuilder is required for an in-housecapability.

APPENDIX C

SUMMARY OF LITERATURE SEARCH ANDDIRECT MAIL CORRESPONDENCE

A thorough review of the photogrammetric literatureand a concerted direct mail campaign were conductedwith the objective of determining past and present uses ofphotogrammetry within shipbuilding and industries hav-ing similar industrial functions. The results are organizedunder two headings: literature and correspondence. Allabstracts of literature have been prepared especially forthis report and are not those of the original authors of thearticles. Pertinent excerpts are included from noteworthyportions of the correspondence.

I. LITERATURE

A. General

1. “Applications of Industrial Photogrammetryin the Shipbuilding Industry" by H. Wienert,Seewirtschaft, 1 (1969), pp. 792-795 (British ShipResearch Association Translation No. 3245).

Following a general introduction to photo-grammetry, the author states that since photo-grammetry was brought into service in the In-stitut fur Schiffbau, Restock only a relativelyshort time ago, described areas of use and pos-sible applications so far recognized or still to bestudied are not necessarily all inclusive. Areasof use are classified as:

a. extending known conventional tech-niques such as measuring propellers, scalemodels, etc.,

b. exploiting new measuring procedureshitherto not possible or at least not econom-ically feasible, and

c. collecting and processing primary data,such as from a model design, for further usein automated design systems and productionprocesses.

Specific applications mentioned include pro-peller dimensioning, checking the accuracy ofstructural parts, dimensioning from models,determination of wave patterns in a modelbasin, determination of wave patterns about aship, predicting the fit of adjoining prefabri-cated sections, mapping of yard and port instal-lations and measuring a ship’s form in prepara-tion for repair of collision damage.

B. Design & Planning

1. “Photogrammetry as an Aid to Manufactureof Ship Piping” by W. G. Smith, British ShipResearch Association Report NS306, 1971.

When design modeling is utilized it is necessaryto extract dimensional information regardingpipe lengths and positions from the model. Atpresent this is accomplished by direct mea-surement of the model with a rule followed bythe preparation of isometric sketches and thenthe pipe arrangement drawings. Because directmeasurement of the model is laborious, often-times inaccurate and the need to prepare thearrangement drawings duplicates the informa-tion inherent in the model itself, a simplisticphotogrammetric method for extracting theneeded information from the model has beendeveloped. Due to inaccuracies in the modelitself, differences between the steelwork de-sign and the actual ship and in the photo-grammetric measurements, it is still expectedthat pipe sketches for closing lengths andmade-to-place pipe will be required. Nonethe-less, it is believed that over 80% of the pipeswithin a machinery space can be handled by thephotogrammetric method. To investigate thepracticality of a simple photogrammetricmethod, a number of photographs were taken

model of a dredger. Measurements were found

2. “Photogrammetry Applied to Pipe Systems ofChemical Plant,” by R. Ferrand, Photogrammet-ric Record, October 1965.

The design of pipe routes is usually accom-plished directly on a scale model which is mod-ular in construction to permit access to centralareas. Pipe centerlines and diameters are represented by the wire and disc technique. Ulti-mately it is necessary to transfer dimensionaldata inherent in the model to diagrams for thepipe fabricator and the plant erectors. Photo-grammetry was seen as a precise method forextracting the necessary information. A par-ticular attraction was that it would eliminatecumbersome manual measurements which of-tentimes require reference to columns or adatum which are not readily accessible.An experiment was conducted to compare thetime required, accuracy and completeness of

1Although not stated by the author, it is believed this is a tolerance.

c-1

the manual and photogrammetric methods. Itwas determined that the photogrammetricmethod generated more data in (at least) lessthan one half the time per dimension, withoutthe blunders which oftentimes occurred withmanual measurements. The maximum discrep-ancy at the scale of the actual plant was lessthan 2 inches by the photogrammetric method.Encouraged by results of initial experiments, aphotogrammetric camera system and a stereoplotter which would plot plan and elevationviews simultaneously were specially devel-oped for dimensioning from design models. Atthe time of preparing the article, a project usingthe new technique had not yet been completed.

3. “Pipe Production Information: AComputer-Aided Method” by B. Dodd and J. B.Jack, British Ship Research Association ReportNS396, 1974.

A simplified photogrammetric process is ap-plied to photographs of a wire and disc machin-ery space model to generate three-dimensionalcoordinates along the course of each pipe (seearticle 1 above). These data are stored directlyin a mini-computer controlled interactivegraphics system which simultaneously dis-plays each digitized pipe in an isometric view.Special programs to generate and displaypipe-bending data also reside within the mini-computer.

4. “Using Photogrammetry to Facilitate Plan-ning and Designing Jobs,” by Josef Paulus, JenaReview 1975/2. ”

When planning new plants or modernizing ex-isting ones dimensioned drawings of machin-ery are required at the scale of the installationplan. Documentation normally provided by thevendor is oftentimes inadequate for this pur-pose. Analog photogrammetry was adopted asan efficient and accurate method for generatingthe necessary drawings. Through the use ofextra photographs extreme reaches of movingand removable elements of a machine are alsoaccurately shown on the drawings. It is alsosuggested that vendors could use the same pro-cess to generate such drawings prior to deliv-ery. The possibility of dimensioning from scalemodels is also mentioned.

C. Erection

1.tures by Photogrammetry,” by J. E. Holliday. J.B. Jack and W. G. Smith, British Ship ResearchAssociation Research Report NS 336, 1971.

In an experiment designed to demonstrate thepotential usefulness of photogrammetry foranalyzing the dimensional equivalence of mat-ing faces of half ships which will be joinedafloat. a partially erected midship section of a135,000-DWT tanker was photographed on itsbuilding berth. A horizontal strip of 11 over-lapping pictures was obtained at a distance of12 meters from the section by moving a singlecamera at specific intervals across special scaf-folding erected for the purpose.

To provide control for the photogrammetricsolution, the locations of ten chalked crosseson the transverse section were determinedthrough conventional triangulation by observ-ing horizontal and vertical angles to the crossesfrom the ends of a baseline which was estab-lished perpendicular to the centerline plane. Toprovide detail within the photographs in loca-tions suitable for correlating adjacent over-lapping photographs, tennis balls were strungon taut piano wires at predetermined locationsrelative to the transverse section. Each stereopair of photographs was set up in a stereo plot-ter and observed data were ultimately trans-formed to the “chalk” control points by thesemi-analytical process. Offsets and elevationsof “hard” points were determined with a toler-ance of approximately 0.11 inch. It is estimatedthat analyses of an entire midship sectionwould require ten days in the shipyard plusanother ten days for data reduction.

2. “Dimensional Quality Control of Large ShipStructures by Photogrammetry” by I. Newton,Photogrammetric Record, October 1974.

In a cooperative effort between the Universityof Newcastle upon Tyne and the Wallsend Re-search Station of the British Ship ResearchAssociation, mating faces of the bow and stemsections of a 258,000-DWT tanker were dimen-sioned by photogrammetnc methods. Ten-sioned piano wires were strung horizontallyand vertically across each mating face so as tomark desired limits of photographic coverage.The camera was operated from a platform

c-2

slung from a crane and the platform was man-euvered so that two horizontal strips of over-lapping photographs could be secured from adistance of 16 meters of each face. Each stripconsisted of 13 photographs. The photographswere taken on rest weekends just prior tolaunchings of the sections and after the sec-tions had been trimmed using a laser device todefine the trim lines. For each transverse sec-tion, work in the shipyard was completed inthree days.To provide control for the photogrammetricsolution a baseline was established somewhatremoved from each transverse section. Fromthe ends of the baseline horizontal and verticalangles were observed to 27 well distributedtargets which were fixed to each transversesection by strong magnets or stenciled withpaint. In addition, elevations of the ends of thebaselines were determined as well as eleva-tions of control point targets along the “mold-ed” edge of the bottom shell.

All stereo pairs were set up in a stereo plotter;approximately 200 detail points on each trans-verse section were observed. XY coordinates(obtained through the semi-analytical process)of the detail points plus the design coordinatesfor the same points were plotted on an automa-tic drafting machine at a scale of 1:100 withdifferences between the design and photo-grammetrically determined coordinates exag-gerated by a factor of 10. Plots prepared for theadjoining faces were superimposed to revealdifferences between the two sections. The tol-erance of determining the detail points is esti-mated to be 0.47 inch in offset and in elevation.

3. “Close-Range Photogrammetry as an Aid toMeasurement of Marine Structures” by IanNewton, Photogrammetric Engineering, De-cember 1975.

Development of the North Sea oil fields hascreated an unprecedented demand in Europefor fixed-base offshore platforms. Typical plat-forms contain some 17,000 tons of steel. areapproximately 500 feet high and measureabout 295 feet by 245 feet at the base reducingto 130 feet by 115 feet at the deck level. Owingto their size, weight and complexity, fabrica-tion of these structures created a number ofproblems. One was to devise an efficient andaccurate dimensional control system to ensurea fit between components when they are liftedinto position during erection. Brown and

c-3

D.

Root-Wimpey Highlands Fabricators Ltd.commissioned a feasibility study to investigatethe potential usefulness of photogrammetry asa dimensional control system for nodes, sub-assemblies and deck stabbing guides. Subse-quent to this study the firm established an in-house photogrammetric capability.

The nodes, up to 50 feet in length, were mea-sured after fabrication to assure that they werewithin specified dimensional tolerances. Dataof interest included radii and roundness of themain tube and all stubs, the direction cosines ofall stubs and the lengths of the stubs. Subas-semblies were measured to assure that theywere within given tolerances for overall sizeand to allow lineup of pile and conductorguides. In addition to overall dimensions,coordinates of the centers of all pile and con-ductor guides were determined so that themeasured unit could be digitally translated androtated into its eventual position on the mainstructure so as to check its fit prior to lifting.Measurement of the relative locations at thedeck stabing guides was performed to deter-mine whether the deck module required ad-justment ashore before its fit was attempted atsea.

The remainder of this article describes sepa-rate work on dimensioning a bulbous-bow unitand mating faces of a tanker built in halves.This latter work is treated in preceding articles.

Propellers

1. “Photogrammetric Measurement of a 26-FootDiameter Propulsion Screw,” Final Report Pre-pared for Aires Marine Shipping Company andNaval Ship Research and Development Center,by DBA Systems, Inc., Melbourne, Florida,January 1975.

After having been in service it was desired todetermine the shape of a 26-foot diameterscrew. Approximately 120 targets attached tothe blades of the screw were triangulated by thefully analytical method using five convergentphotographs. Tolerances achieved were about0.010 inch in each coordinate direction. Sev-eral targets on the hub were used to define aplane from which targets on the blades werereferenced.

2. “The Use of Industrial Photogrammetry in theProduction of Ships Screws” by Gunter Knodlerand Helmut Kupke, Jena Review 3/1974.

A photogrammetric technique for measuringraw and finished propellers during the runningproduction process is being developed jointlyby VEB Dieselmotorenwerk Rostock, theRestock Institute for Shipbuilding and the JenaOptical Works. The method is first used todetermine distortion in and required correc-tions to the mold. In the second phase photo-grammetry is used to generate a digital modelof the casting. The digital model is computercompared to the design to determine necessarymachining refinements. In the last phase thefinished propellers again photogrammetricallymeasured and computer processed in the samemanner as the casting.

3. “A Photogrammetric Method for Measure-ment of the Thickness of the Cavitation Stratumon a Model of a Ship Propeller, Conducted in oneof the Ship Model Tank’s Cavitation Tunnels” bySvein Olav Tveitdal, The Norwegian College ofEngineering, Institute of Geodesy and Photo-

grammetry, January 1972.

This thesis work gives complete details ondevelopment of a photogrammetric system formeasuring the extent and thickness of the cavi-tation stratum about a model of a propeller.Topics treated include calibration of the cam-era system (two electric Hasselblad 500cameras), calibration of the photogrammetricstereo plotter (which was actually used as acomparator), measurement corrections re-quired due to water-glass-air interfaces andphotogrammetric calculations.

After experimenting with a dummy speci-men photogrammetric measurements weremade of a model of a propeller rotating in acavitation tunnel. Prior to inserting the propel-ler a series of targets were adhered to the pro-peller blades and the relative locations of thetargets were measured with the aid of a threeaxis milling machine. The two cameras and astrobe unit were mounted outside of the tunnelviewing through a specially inserted planeparallel glass viewing port. Fourteen stereopairs of photographs corresponding to sevenpositions of the propeller in each of two cavi-tation states were subjected to the photogram-metric analysis. Photogrammetric measure-ments were taken on the targeted points and onthe apparent surface of the cavitation stratum.The former points allowed transformation ofthe in-water data to the free-air condition sincethe locations of the targets had been previously

c-4

measured on the milling machine. Once themathematical transformation was determined,measurements on the cavitation surface werethen transformed to the free-air condition.Differences between these transformed dataand the previously determined (by milling ma-chine measurement) shape of a blade gave thethickness of the cavitation stratum. Data wereextracted in radial sections through the stratumwith an estimated tolerance of determining

4. “On Stability and Volume of Marine PropellerCavitation and Corresponding Spectral Distribu-tion in Hull Pressure Fields” by K. Holden. T.Sontvedt and O. Ofsti, Symposium on High Pow-ered Propulsion of Large Ships (Part 1), Publica-tion No. 490, Netherlands Ship Model Basin,December 1974.

Full scale investigations of propeller cavitationnecessitated development of equipment andprocedures to detect blade cavity thickness,time variations of blade cavitation, amplitudesof propeller induced pressure fluctua-tions-stability. phase relationship and spectraldistributions, and natural ventilation of cavita-tion. Cavity thickness distribution is deter-mined by photogrammetric methods andnatural ventilation, with a limited amount of airsuction, is observed by TV and photogram-metry. Photo and flash equipment were placedin steel tubes welded into the hull of a 224.000-DWT tanker. Photogrammetric analyses wereaccording to the methods developed by Tveit-dal as described in article 3 above.

5. “Recent Development of Industrial Photo-grammetry in Japan,” by Dr. Taichi Oshima,Photogrammetric Engineering and Remote Sens-ing, March 1976.

In a review of recent engineering applicationsof photogrammetry, one paragraph briefly de-scribes measurement of the thickness of cavita-tion bubbles about a model of a propeller. Thefully analytical method of photogrammetrywas employed.

E. Model Basin

1. “Wave Surface Configuration” by ProfessorFrancis H. Moffitt, Photogrammetric Engineer-ing, February 1968.

A makeshift photogrammetric system em-ploying a pair of synchronized Rolleiflex pic-torial cameras was developed to measure

waves produced by a model of a ship in a tow-ing basin. Stereo pairs of photographs corres-ponding to different towing velocities and wa-ter depth conditions were observed in a stereoplotter. For deep water runs the graphic outputof the stereo plotter was in the form of profilesalong the crests of the bow and stem wakes,with selected cross sections perpendicular tothe profiles. For all shallow water runs thestereo plotter was used to contour the wakes inintervals of 0.2 inch relative to the still waterlevel.

2. “A Photogrammetric Space-Time-Problem:Locating of Ship Positions for the Determimtionof Test Model Properties” by Dr. techn. F. Los-chner, Bulletin dela Societe Franqaise de Photo-grammetry, No. 41 (in German).

Photogrammetry was used to determinecourse stability, maneuverability and cruisingspeed as a function of engine output and rudderposition for a large test model built according toa new design. The model was fit with strobelights so that the tests could be conducted atnight with the camera shutters always remain-ing open. The trajectory of the model was re-corded by two cameras as a series of flashes ordots on the photographs. Positions of the dotsrepresenting the trajectory were determinedwith an accuracy of 10 cm. 1 Precise timing ofthe flashes also allowed a determination ofspeed.

F. Miscellaneous

1. “Photogrammetric Survey of Trajectory ofLaunched Ship” by A. Linsenbarth and A.Rymarowicz, Presented to the 1972 Congress ofthe International Society of Photogrammetry, Ot-tawa, Canada.

A pair of cameras with shutters synchronizedfor exposures at intervals of 3 seconds wereused to photograph a 26,000-DWT ship as itwas launched. Three points on the ship andseveral stationary control points on the groundwere signalized with lights so that successiveexposures could be taken on one set of photo-graphic plates. Analog photogrammetricevaluation allowed mapping of the three di-mensional trajectory of the ship. Since the timebetween exposures was known. the speed ofthe ship at several points was also determined.

2. “The Photogrammetric Measure of the Speedof the Magdala Tanker Towed in the Harbor ofSaint-Nazaire,” by M. Carbonnelle, Bulletin d’

French).

Magdala is the first French tanker in the200,000-DWT class. To assist alignment dur-ing transfer to an outfit berth, a circular pier ofabout 65 feet in diameter was built approxi-mately 130 feet in front of the berth. This pierwas to function as a pivot and was equippedwith very powerful shock absorbers. Toevaluate the effectiveness of these devices itwas necessary to determine the trajectory andspeed of the ship just prior to - contact, at con-tact, and just after contact. Velocity data wasto be within 1 centimeter/second but absolutepositioning of the trajectory of the ship was notas critical.1 Since conventional surveyingtechniques did not provide a suitable solution,a photogrammetric scheme was devised.Targets were placed on the hull of the ship andalso on land such that all could be seen by twocameras. The camera shutters were syn-chronized and actuated at precise intervals intime. Fully analytical triangulation of thetargets provided information regarding posi-tion of the ship. The time interval betweensuccessive stereo pairs of photographs alloweddetermination of velocity as well.

3. “Recent Development of Industrial Photo-grametry in Japan,” by Dr. Taichi Oshima,Photogrammetric Engineering and Remote Sens-ing, March 1976.

This article provides a review of more recentuses of photogrammetry for industrial pur-poses. Three paragraphs are given to the use offully analytical photogrammetry, in combina-tion with laser measurements, 35-mm movingpictures and soil pressure measurements forthe purpose of monitoring subsidence of thelaunching ways during the launch of a largetanker. Measurements were performed at theMitsubishi dock by the Ohbayaski Construc-tion Company.

4. “Instant Draughtsman in the Dockyard,” byAngela Croome, New Scientist, August 1972.

Photogrammetry is playing a major role in thefive-year plan for refurbishing the external

1 It is not known whether this is a tolerance or one standard deviation.

c-5

stem section and redecoration of the cabinswithin the stem of the warship HMS Victory,Nelson’s flagship at the Battle of Trafalgar.Photogrammetrically prepared detailed draw-ings of the complex stem section and of thecabins are used by the shipwright craftsmen tofashion wood refurbishments.

1. Letter of 4 April 1975 from Professor KennertTorlegard, Royal Institute of Technology, Divi-sion of Photogrammetry, Stockholm, Sweden tojohn f. kenefick, Photogrammetric Consultant,Inc.

During my time as consulting engineerwith VIAK AB, we performed tests withphotogrammetric technique in order to, for ex-ample:

-check the shape of steel plates for the bulb,bow and stem of ships-measure ship sections being built in sepa-rate wharfs and predict the disclosures inthe assembling work-check the final shape of the hull-digitizing the shape of wooden car models-check the dimensions and shape of pre-fab-ricated housebuilding elements

In many cases the above measuring problem was better solved by other techniques thanphotogrammetry, due to the time requirementsand the wishes of an in-house method.”

2. Letter of 2 May 1974 from Dr. Ing. Walter S.Ferri, Officine Galileo, Florence, Italy to john f.kenefick, Photogrammetric Consultant, Inc.

"The picture sent to you shows an oil tanker onthe whole (about 300,000 tons) in a shipyard.We were required for the control of faces A andB of the two contiguous pieces,1 but the photo-grammetric solution has not been found.”

3. Letter of 25 February 1975 from AssociateProfessor Wolfgang Faig, The University of NewBrunswick to john f. kenefick, Photogram-metric Consultant. Inc.

“Unfortunately, your letter of Feb. 17 request-ing information on Photogrammetric Applica-tions in the shipbuilding industry arrived abouttwo days too late. Last Thursday, we had a

guest lecturer here at UNB addressing exactlythis very problem. Dr. Aleksandra Bu-jakiewica, professor at the Warsaw Polytech-nique is just completing her term as a visitingscientist at the National Research Council inOttawa. She was a member of a Polish studygroup working on this problem at the shipyardsin Gdansk, Poland. ”

4. Verbal discussion at General Dynamics Cor-poration, Quincy Shipbuilding Division disclosedthat the firm Bloms Oppmaling A/S of Oslo,Norway has patented a combined survey-ing/photogrammetric technique for developingsounding tables for LNG spheres. Volumes areguaranteed to be within O. 1%.2

1 It is believed this is in reference to predicting fit-up between separately-const2It is not known whether this is a tolerance or one standard deviation. ●

rutted bow and stem sections.

C-6

APPENDIX D

SOURCES OF PHOTOGRAMMETRIC SERVICES

Most photogrammetric work relates to aerial mappingwhich compared to hind-based short-range applications isvirtually the entire U.S. market for photogrammetric ser-vices. Thus, most photogrammetric firms deal primarilywith aerial surveying as of March 1976 there were onlytwo known to specialize in short-range work.

While it is true that aerial-survey firms could conceiv-ably work with land-based photography, it is a general rulethat most aerial survey firms lack experience in the kind ofplanning and implementation that is prerequisite for indus-trial processes. Moreover, these firms may tend to applyequipment and software, designed for aerial use, whichmay not be suited to shipbuilding. This is not to say thatthere are no aerial survey firms having experience withland-based projects. Those that are known to have someexperience, even though limited, are listed herein.

It is strongly recommended that a shipbuilder seekingphotogrammetric services require that each respondentsubmit the following information for evaluation prior toany subcontract award:

a. Pertinent recent experience using land basedcameras. including description of project, equip-ment used, method(s) of photogrammetry em-ployed, results achieved (including accuracytests) and names, addresses and telephone num-

b.

c.

hers of personnel familiar with the work withinthe customer’s organization.

Planned method of operation for the shipbuilder’sproject, including equipment to be used,method(s) of photogrammetry to be employed, anaccuracy analysis which demonstrates that theshipbuilder’s tolerances can be achieved, aschedule of performance and identification of allproposed second-tier subcontractors to be em-ployed by the photogrammetric firm.

Discussion of corporate background, most recentfinancial statement(s) covering 12 months of op-eration and resumes of key personnel to bespecifically assigned to the project.

The shipbuilder must be fairly specific as to data re-quired and its method of presentation. Although, this isoftentimes refined by mutual agreement. Above all, theshipbuilder must be firm on tolerances required of the dataand, as already indicated, the photogrammetrist must beable to demonstrate that his methods can produce datawhich conform to these demands. This is a point of cau-tion. Photogrammetrists have been known to become in-volved with just finding a solution at the expense ofsufficiently accurate results.

On the other hand, the shipbuilder must not overstaterequirements as this could lead to a noncompetitive situa-tion. Also, data derived under very rigid specificationsmay be of no greater practical value than that from realis-tically stated requirements.

See pages D-2, D-3 and D-4 for service, consulting, andhardware firms.

D-1

SERVICE FIRMS1

circa March 1976

FIRM SERVICES MARKETED

john f. kenefickPhotogrammetric Consultant, Inc.z

P.O. BOX 3556Indialantic, Florida 32903(305) 723-8515

Henderson Aerial Surveys, Inc.5125 West Broad StreetColumbus, Ohio 43228(614) 878-3925

Raytheon Company/Autometric400 Army-Navy DriveArlington, Virginia 22202(703) 979-6100

LaFave, Huntley. White andMcGivern

850 Hudson AvenueRochester, New York 14621(716)467-1010

“DBA Systems, Inc.4

P.O. Drawer 550Melbourne, Florida 32901(305) 727-0660

Ralph L. Woolpert Company2324 Stanley AvenueDayton, Ohio 45404(513) 461-5660

computer programmingfully analytical photogrammetryanalog and semi-analyticalphotogrammetry 3

analog photogrammetrysemi-analytical photogrammetry

computer programminganalog photogrammetryfully analytical photogrammetry

analog photogrammetry

computer programmingfully analytical photogrammetry

analog photogrammetrysemi-analytical photogrammetry

1This listing includes firms believed to have fairly recent pertinent experiences in the indicated areas. However. in several instances the depth of theseexperiences is not known and except for the first listed the experiences are not known to be within shipbuilding. Additional information may be solicitedfrom the American Society of Photogrammetry. 105 North Virginia Avenue. Falls Church. Virginia 22046: (703) 534-6617.2Specializes in and actively seeks out land-based photogrammetric work.3stereo plotter work is subcontracted.4Specializes in and actively seeks out land-based photogrammetric work.

D-2

CONSULTING FIRMS1

circa March 1976

john f. kenefickPhotogrammetric Consultant, Inc.P.O. BOX 3556Indialantic, Florida 32903(305) 723-8515

Donald R. Graff, P.E.Consultant in Surveying and MappingP. O. BOX 311Beaver Dam, Wisconsin 53916(414) 885-9191

LaFave, Huntley, White and McGivern850 Hudson AvenueRochester, New York 14621(716) 467-1010

D-3

HARDWARE MANUFACTURERS

FIRM PRODUCTS MARKETED

Wild Heerbrugg Instruments, Inc. comparators465 Smith Street stereo plottersFarmingdale, Long Island, New York 11735 cameras(516) 293-7400

DBA Systems, Inc. comparatorsP. O. Drawer 550 cameras (special order)Melbourne, Florida 32901(305) 727-0660

Gallileo Corporation of America comparators 150 Fifth Avenue stereo plottersPelham, New York 10803 cameras(914) 738-1154

Zena CompanyP. O. BOX 338South PIainfield, New Jersey 07080(201) 754-4109

Kern. Instruments, Inc.111 Bowman AvenuePort Chester, New York 10573(914) 939-0200

O.M. I. Corporation of America1319 Powhatan StreetAlexandria, Virginia 22314(703) 549-4064

Carl Zeiss, Inc.444 Fifth AvenueNew York, New York 10018(212) 730-4400

Danko Arlington, Inc.Kelsh Instrument Division4800 East Wabash AvenueBaltimore, Maryland 21215(301) 664-8930

stereo plotterscameras

comparatorsstereo plotters

comparatorsstereo plotters

comparators

comparatorsstereo plotterscameras

cameras (special order)