JAMES H . KING & MATTHEW D.

DEVINE

HANDE - A COMPUTER-AIDED DESIGN

APPROACH
FOR HYDROFOIL SHIPS

L
EDITOR’S NOTE: For the discussion of the sixfeen Computational Program Capabilities, the attention
of the reader is invited to the APPENDIX to this paper.
1
THE AUTHORS program for the purpose of increasing the ship
designer’s productivity. Those few software projects
Mr. James H. King graduated from Webb Institute of Naval that have attempted an integrated ship design capability
Architecture in 197.5 at which time he received his B.S. degree have resulted in still fewer satisfactory efforts.
in Naval Architecture and Marine Engineering. He joined the
David W . Taylor Naval Ship Research and Development One major software project that successfully achieved
Center (DTNSRDC) following graduation where he has work- an integration of design and analysis capability for
ed in both the Advanced Concepts and Hydrofoil Offices. hydrofoil ships culminated in the U.S. NAVYHydrofoil
Currently, he is the Assistant Manager of Hydrofoil Systems ANalaysis and DEsign (HANDE) Computer Program.
Integration in the Advanced Hydrofoil Systems Office at HANDE was designed to avoid the pitfalls typical of
DTNSRDC. Besides ASNE, which he joined in 1977, Mr. programs of similar scope, such as extreme difficulty of
King is an Associate Member of SNAME and a member of the use, poor responsiveness to engineering queries, and in-
U.S. Naval Institute, and the International Hydrofoil Society. adequate technical depth in the multi-disciplined en-
vironment. The HANDE engineering tools for ship
Mr. Matthew D. Devine attended the University of Michigan design are manipulated by the user via a small, yet
from which he received both his B. S . E. and M . S.E. degrees in
Naval Architecture and Marine Engineering in 1973 and 1974
powerful, set of commands. HANDE was designed to
respectively. FoNo wing graduation, he joined the Boeing execute interactively via a teleterminal to provide desk-
Company where he has worked in computer-aided ship design top convenience while avoiding delays inherent in batch
and analysis and is currently employed in the Space and (card) oriented systems. Finally, HANDE incorporates
Military Applications Division, Boeing Computer Services virtually all major technologies that are relevant to
Company, Seattle, Wash. He is an Associate Member of hydrofoil-ship design. HANDE has consequently allow-
SNAME and has been a member of ASNE since 1980. ed a dramatic increase in engineering productivity dur-
ing the hydrofoil-ship design cycle by allowing the user
ABSTRACT to apply the HANDE engineering tools in an easily-
used, responsive, yet technically sophisticated
A powerful computer-aided design tool for use in hydrofoil environment.
ship engineering, the Hydrofoil ANalysis and DEsign program Use of the HANDE engineering system closely
(HANDE), is described. Its relev’ance, structure, features, and
parallels the classical process of ship design. The design
use are delineated. The value of HANDE for design verifica-
tion and variation, research studies, and rapid response team begins with a set of mission requirements that the
studies is related through case histories. Future application proposed ship is t o accomplish. Existing design data and
and development of HANDE and related design tools are computational procedures are employed in an iterative
forecast. sequence t o derive a ship design, as exemplified by the
design spiral by MILLER[ l ] shown in Figure 1.
INTRODUCTION HANDE’s value is in the automation of many of the
manual processes performed in the iterative design pro-
T H E ADVENT OF T H E MODERN D I G I T A L COMPUTER has cess. Instead of manual search through lengthy tables of
tremendously altered computational efforts within all residuary resistance coefficients, HANDE performs the
fields of engineering. The domain of the naval architect search. Instead of manual construction of a plot of
has been no less affected. Many large computer soft- hydrostatic righting arm versus heel angle, HANDE
ware projects have resulted in engineering tools draws the plot. Instead of manual storage of design data
(programs) that have expanded and enhanced ship in dust-covered notebooks, HANDE stores the data on
design capability. Most of these tools have been con- computer disk files from where it may be easily recalled
cerned with a single discipline of calculations such as and reviewed.
hydrostatics, resistance, strength, propulsion, or Although many of the processes involved in the
seakeeping. Unfortunately, very few marine-related design of a ship are automated by HANDE, the pro-
software projects have attempted the integration of gram leaves the critical engineering decisions to the
several ship-design disciplines into a single computer designer. HANDE makes no attempt to decide whether
120 Naval Engineers Journal, April 1981
KINWDEVINE HANDE/COMPUTER-AIDED DESIGN APPROACH

GENCRILARRANGLMLNT ture, command language, and computational capabil-

R VOLUMES
ities will be described. A simplified case of HANDE
usage will be demonstrated. HANDE’s relevance in an
engineering environment will be related through case
histories of its use in design verification and variation
tasks, research efforts, and rapid-response studies. Its
LENGTH WEIGHTS
future development and application will also be
projected.

HISTORICAL
DEVELOPMENT
D R A F T FHL L tl UAHU ENDURANCE
DEPTH A N D R L A h l OIL

Development of the HANDE Computer Program

formally began in December 1972 with an award to the
HULL FORM RESISTANCE
Boeing Company of a Phase 0 development contract
k
PROPULSION
from the David W. Taylor Naval Ship Research and
UISPLACEMCNT Development Center (DTNSRDC). Under Phase 0, the
technical and software specifications for development
Figure 1. The Design Spiral _of HANDE were to be clearly and explicitly defined.
HANDE Phase 0 concluded with the publication of a
to employ waterjet or propeller propulsion, whether to final report by Brennan, Burroughs, Hurt, Wichert, and
use Newton-Radar or Wageningen-B propeller curves, Wacher in June 1973 [2].
or whether to use a three- or four-bladed propeller. Construction of the HANDE computer software was
HANDE makes no significant design decisions initiated following the close of HANDE Phase 0. The
whatsoever. The program simply employs selected bulk of software construction occurred during HANDE
algorithms to perform selected calculations. The Phase 1, which commenced in March 1975. Several
designer retains essential control of the program. preliminary program versions were made operational on
In the sections that follow, HANDE’s historical the Boeing Computer Services Company EKS Com-
development, its conceptual organization, data struc- puter installation which features Control Data Corpora-

IN IT I AL I ZAT I ON

SYNTHESIS
-
ANALYSIS

COMPUTATIONAL PROGRAMS
-

Figure 2. HANDE Program Conceptual Organization

Naval Engineers Journal, April 1981 121
HANDEICOMPUTER-AIDED DESIGN APPROACH KING/DEVINE

tion Cyber mainframe equipment. The first fully opera- ship data may be transferred from the data bank to the
tional HANDE Program, version 0.0, was assembled current model or from the current model to the data
and delivered to the NAVYin April 1977. Complete bank.
program documentation was delivered at the close of The execution program, the current model, and the
HANDE Phase 1 in June 1977. data bank can also be employed t o create entirely new
Subsequent HANDE development work has expand- ships in the current model by recall from the data bank
ed program capability and improved performance. The of pieces of data from different ships. For example, one
current version, HANDE 1.3, consists of more than can transfer ship data corresponding to the propulsion
54,000 lines of FORTRAN computer code. system of Ship A from the data bank to the current
model and then transfer hull offsets for Ship B from the
P R O G R A M CONCEPTUAL ORGANIZATION data bank t o the current model. The current model
would thereby contain a vessel of Ship A type propul-
The HANDE System is composed of five principle sion system but Ship B type hull.
elements: 1) T H E DESIGNER, 2) A N EXECUTIVE PROGRAM, 3) A
SERIES OF COMPUTATIONAL PROGRAMS, 4) A SHIP DESIGN Computational Programs
U N D E R G O I N G GENERATION OR ANALYSIS (called “current
model”), and 5) a DATA B A N K . These elements are shown The calculative function of HANDE is performed by
in Figure 2. the element that consists of sixteen computational pro-
grams [See APPENDIX]. Each program represents a
Designer distinct engineering technology that can be applied to
the design and analysis of hydrofoil ships. Through a
The designer is the controlling element of the simple command t o the executive program, any one of
HANDE system. Through a simple command language, these programs may be executed. All input data re-
the designer directs execution of, or interaction be- quired by the given computational program is
tween, the remaining system elements. Although automatically taken from the current model by the ex-
capable of batch (via cards) execution, the HANDE ecutive program and given to the computational pro-
system was designed as an interactive tool for hydrofoil gram. Following termination of the computational pro-
ship engineering. Consequently, the designer typically gram, output data, selectable by menu, are displayed to
utilizes HANDE by means of a tele-terminal where com- the designer. Certain computational programs also add
mands may be entered and results of those commands to, or modify, the current model as part of the ship
immediately reviewed. Delays are thereby minimized. design-generation process.

Executive Program DATASTRUCTURE

The executive program is the HANDE system element Model Parameter List
that interprets each user command and thereafter per-
forms each task that is required to accomplish the user Central to use of the HANDE design system is an
instructions. The executive program is also the lone understanding of the data elements contained within a
system element that can interact directly with each of current model and the organization of those data
the other system elements. Performance of any given elements. Overwhelming amounts of ship data, ranging
user command generally involves the remaining three from hull offsets to number of anchor chain links, could
system elements. theoretically be placed in a current model. However,
much ship data has limited relevance in early design
Current Model work. Furthermore, excessive amounts of data tend to
obscure the important and highly critical data during
The current model element of the HANDE system is early design phases. To insure that the current model
the temporal collection of data that represents the one does not contain superfluous or unnecessary data, only
hydrofoil ship design under generation or undergoing data that are input to one or more of the given computa-
analysis. All program computations use current model tional programs are contained within the current model.
data only. The current model is temporal because it ex- Organization of the current model is accomplished by
ists only during execution of the program. T o become means of a four-tier, tree-type hierarchy, known as the
permanent, current model data must be transferred to a Model Parameter List (MPL). This structure was
permanent storage device. selected to provide an adequate organizational breadth
for the data yet retain a sufficiently simple structure to
Data Bank avoid confusion in assemblage, storage, or recall of cur-
rent model data.
A data bank has been incorporated as an element of The highest tier of the MPL represents the entire ship
the HANDE system for the purpose of permanent reten- system. The next tier consists of groups, which are fur-
tion of ship data. Entire current models or pieces of a ther subdivided into sub-groups. Finally, the sub-groups
current model may be stored in the data bank under a are each divided into the parameters which are the
name selectable by the user. During a HANDE session, lowest tier of the hierarchy. No specific data are
122 Naval Engineers Journal, April 1981
KINWDEVINE HANDE/COMPUTER-AIDED DESIGN APPROACH

LEVEL
-

SYSTEM

4--+
GROUP

r
FB E N G I N E FB GEARBOX FB PROPELLER F B WATERJET SUB-GROUP

1
1 21.800.0 HP 3000 R P M 0.40 L B M / H P - H R PARAMETER
FB NO E N G FB CONT HP AVAIL F B CONT RPM F B SFC

Figure 3. HANDE Model Parameter List Hierarchy

associated with the ship, group, or sub-group tiers. Data Types
They exist as a means of simply organizing the actual
data elements associated with the parameters. Wherever Parameters may represent any of four data types:
feasible, the NAVYShip Work Breakdown Structure SCALAR, VECTOR, ARRAY, or NON-NUMERIC. Non-numeric
(SWBS) has been employed in subdivision of the MPL parameters are termed “indicators” and are used as
into groups and subgroups. decision points by the computational programs. For ex-
Figure 3 shows an abbreviated model parameter list ample, the parameter representing the foilborne propul-
hierarchy utilized by HANDE. The system level sion type indicator may have either of the values PRO-
represents the entire ship. The groups and sub-groups PELLER or WATERJET, corresponding to propeller or
represent either major physical subdivisions of the ship, waterjet type propulsion respectively. Such a parameter
such as the hull, foilborne propulsion system, and for- is used, for example, by the designer to order the com-
ward foil/strut groups, or major areas of analytical con- putational program responsible for calculation of
cern, such as the ship mission requirements, cost fac- foilborne propulsion data to use propeller-propulsion
tors, and performance groups. algorithims and data instead of those of waterjet
Division of a group into sub-groups is exemplified by propulsion.
the foilborne propulsion group, as shown in Figure 3.
Sub-groups to the foilborne propulsion group include EXECUTIVE
COMMANDS
foilborne engine, foilborne gearbox, foilborne pro-
peller, and foilborne waterjet. Finally, the actual data Form
reside in the parameter tier of the MPL hierarchy.
Parameters within the foilborne engine sub-group are Commands are issued by the designer to the HANDE
number of engines, continuous horsepower available, Executive Program in the form of brief command
continuous RPM, et cetera. As shown in Figure 3, each strings. Each command string is composed of one or
of these parameters has a specific value associated with more phrases, and each phrase is no greater than twenty
it. characters in length. A phrase is separated from other
Naval Engineers Journal, April 1981 123
HANDE/COMPUTER-AIDED DESIGN APPROACH KINWDEVINE

phases by a delimiter, such as a comma or equal sign. COMPUTATIONS

The Executive Program reads one command string at a
time (corresponding to one line at the teleterminal or Computational Programs
one card in a batch-type run) and then breaks the string
into its distinct phrases. Beginning with the first phrase, Three types of computational programs exist within
each phrase is diagnosed as to the command it implies, HANDE: lNlTlALlZATlON, SYNTHESIS, and ANALYSIS. The
and the program performs the steps necessary to per- breakdown of programs within each type is shown in
form each command. Figure 4. A summary of the computational functions
performed by each program is given in the APPENDIX.
The initialization section of HANDE consists of a
single program. It utilizes simple empirical methods to
Three types of commands are recognized by the calculate a variety of ship data. As its name implies, a
HANDE Executive Program: 1) current model/designer primary function of the initialization program is to pro-
data flow commands, 2) program call commands, and vide an initial starting point for a new design under
3) general executive commands. Each of the three types development with HANDE. A secondary use of the in-
of commands are explained in the following sections. itialization program is in performance of high-level,
parametric trade studies.
Ten synthesis-type computational programs, shown
CURRENT MODEI./DESIGNER DATAFLOWC O M M A N D - in Figure 4,exist within HANDE. Each program is con-
This command type permits the designer to enter data cerned with a single technological area of hydrofoil-ship
manually into the current model or to have current design. In contrast to the initialization program, each
model data output to the designer for examination. T o synthesis program utilizes rigorous analytical techniques
enter data into the current model, the name of the in computation of ship data.
parameter for which data is to be specified is typed as The third type of computational program is called
the first phrase of the command string followed by the analysis, of which there arefive. Like the synthesis pro-
value the parameter is to assume. Data entry for a scalar grams, rigorous analytical techniques are employed.
parameter is performed in this manner. Data entry for The principal difference between synthesis programs
vector, array, or indicator type parameters is similar. and analysis programs is that synthesis programs
T o examine the data value in the current model of any modify the current model. Analysis programs do nor
given parameter, the designer simply enters the name of
the parameter as a command string. The data value cor-
responding to the specified parameter would then be
printed for examination by the designer.

SYNTHESIS
PROGRAM CA1.i. COMMANDS - The program call com-
mand causes a specific computational module to be ex-
ecuted, using input data from the current model. The
program call command consists of the name of the com-
putational program to be executed.

G E N E R AE
L X E C U T I VCOMMANDS
E - Thirty-three
general executive commands exist which enable the
designer to perform a variety of tasks. Among the func-
tions that can be performed with these commands are:

Initiation and control of a process known as syn-

thesis, which is an automated execution sequence
of several selected computational programs.
Selection of output t o be generated upon execution
of any given computational program.
Data transfer between the current model and data
bank.
CLND
Scan of current model for missing data.
Listing of data bank contents, available com-
mands, available computational programs, current
model parameters, and other information.
Inputloutput unit selection (metric or English
units) .
Miscellaneous functions. Figure 4. HANDE Computational Modules

124 Naval Engineers Journal, April 1981

KING/DEVINE HANDE/COMPUTER-AIDED DESIGN APPROACH

modify the current model and provide little more than TABLE 1
additional information about it. Also unlike analysis
programs, synthesis programs can be employed in an OUTPUT MENUS FOR PERFORMANCE ANALYSIS
iterative loop to converge on a ship design. This process, PROGRAM
known as design synthesis, is simply an automated
traverse of a design spiral from the mission re-
quirements t o the converged ship design. Performance Analysis Summary
Performance Analysis Range Factor
Design and Range Point Data
Design Synthesis Design and Range Point Power Data
Design and Range Point Propulsion Data
The design synthesis process is another step employed GRAPHICS
MLNUNo Tiii~
in the manual process of ship design that has been incor-
porated into HANDE. After establishment of mission Drag vs. Speed
Range vs. Speed
requirements, the designer typically generates an initial
Power vs. Speed
design to serve as a starting point. This initial design Fuel Flow vs. Speed
may be a previously established design of similar func- Specific Fuel Consumption vs. Speed
tion or an entirely new concept. Unfortunately, the in- Propulsion Efficiency vs. Speed
itial design is seldom satisfactory. Minor or gross Fuel Consumption vs. Speed
modifications must be performed. For example, addi- Transport Efficiency vs. Speed
tional cargo volume may be needed. The designer may Average L I D vs. Speed
elect to expand the hullform t o satisfy this need. But ex-
panding the hullform changes ship resistance, which im- feeds the data to the program. The Executive Program,
pacts required propulsive power, which may demand a following the orders of the designer, also controls the
new power plant, which may change the amount of fuel output to the designer. All computational programs
carried to achieve a desired range. The modified hull, produce printed output and may also produce graphical
propulsion plant, and available fuel each impact the output. The printed and graphical output for a given
total weight of the ship. The initial estimate of ship computational program are divided into items which are
displacement for which resistance calculations were selectable by menu number. Thus, whenever a computa-
previously performed may require revision, and new tional program is executed, the Executive Program in-
resistance calculations may need to be performed. The sures that only those output items specifically requested
design spiral goes on and on, hopefully toward a con- by the designer are printed or displayed.
verged design. Example menus of output items for a computational
Such is the design synthesis process employed by program are shown in TABLE 1. The menus shown in
HANDE, but with the added complication of hydrofoil TABLE 1 correspond to those for the performance
systems. The iterative scheme employed by the HANDE analysis program. Five print items and nine graphics
design synthesis process, as shown in Figure 4, is com- items exist. Any or all of these items may be selected for
posed of two loops: A N I N N E R LOOP and A N OUTER LOOP. output by the designer.
The inner loop attempts convergence on hydrofoil In addition to output to the designer, the initialization
system characteristics whereas the outer loop attempts and synthesis type programs also output data which the
convergence on the entire ship system. Executive Program uses to update the current model.
The key to operation of the design synthesis process is The update of the current model is performed
the ability of each synthesis module to modify the cur- automatically by the Executive Program without
rent model. Critical ship data in the current model, such designer interface.
as hull lines, superstructure characteristics, foil system
geometry, foilborne drag, foilborne propulsion data, DOCUMENTATION
hullborne drag, hullborne propulsion data, foil system
structural characteristics, fuel/range data, and ship A computer as large and complex as HANDE re-
weights, are modified during the synthesis process; each quires comprehensive documentation. Without it,
by the appropriate computational program. Con- growth of the User Community would be impossible or
vergence of a ship design occurs when two passes severely hindered. HANDE is documented in five
through the outer synthesis loop produce virtually iden- volumes by BRENNAN, BURROUGHS, KNUTSEN, MELDAHL,
tical designs. STRAIN, and WACKER [2] [3] [4](51 [ 6 ] . These volumes
provide a Summary Manual, a User Manual, a Theory
Executive Relationships Manual, a Programmer Manual, and a Data-Bank
Maintenance Manual.
The role of the Executive Program in execution of a
computional program is limited to control of input to, SAMPLE
USE
and output from, the computational program. When a
computational program is to be executed, the Executive An extremely simplified case of HANDE usage will
Program extracts the input data required by the com- now be demonstrated. A ship that has been previously
putational program from the current model, and then designed through use of HANDE will be utilized. The
Naval Engineers Journal, April 1981 125
HANDE/COMPUTER-AIDED DESIGN APPROACH KINWDEVINE

HYDROFOIL ANALYSIS AN0 DESIGN PROGRAM (HPNDE) I> 8 8 / 0 6 / 2 8 . 17.52.16.

VERSION 1.3 H B HVDRODVNRllICS CRRPHICS nENU NO. 1
DATED MARCH 1 7 . 1980
;-,-,
240000.0 I I
CMD.E>LIST COMMANDS
7 PITON
SHIPS GROUPS PARAMETERS 200000.0
USE STORE MODIFY 3 MTON
REMOVE CURRENT MODEL EXIT
DIAGNOSTIC CONTROL L I S T COPMANOS RE I N I T I AL I ZE 9 MTON
ONLINE OFFLINE L I S T MODULES 160000.0
LINEPRINTER BATCH DESIGN
R
IIESICN P R I N T RATE SUMMARIES SKIP A
iii-ioau~is INCLUDE NO MODULES c
DESIGN INTERATIONS I N D STATUS CPU T I M E 120000.0
SCAN 1/0 ENGLISH U N I T S METRIC U N I T S N
GRAPH 1CS BAUD RATE READ DATA

CMD.E>SHI PS 80000.0

MOOEL 908 MDDEL 901 HOC

AGEH JETFOIL MODEL 1 0 2 6
PCH-1 40000.0
TUCUMCARI

CMD.E>USE.MODEL 90B
CMO, E>LEP 0.0
LEP = 118.119 3.0 6.0 9.0 12.0 15.0 18.0 21.0 24.0
CMD.E>METRIC U N I T S SHIP SPEED, K T
CMD ;M>LEP
LBP = 36.000
CMD.lOGRAPHICS.HE HYDRO’ MODULE, 1
CMO,M>HE HYDRO MODULE Figure 6. SPEED vs. DRAG Plot Constructed by HANDE
HE HYDRO M W U L E “FATAL ERROR** NO. 4
I N V A L I D DATA - HB DRAG MDDE I N D
CMD.M>HE DRAG MDDE I N D = PLANING selected for output from this computational program by
CMD,M>HE HYORO MODULE
the GRAPHICS, HB HYDRO MODULE, 1 command.
OlO.M>EXIT
The program is executed by entering the name of the
computational program, HB HYDRO MODULE.
Figure 5. Example HANDE Use As shown in Figure 5 , a fatal error has been detected
design data for this ship are stored in the data bank by the computational program. The error diagnostic in-
under the name MODEL 900. Figures 5 and 6 shown the forms the user that the parameter HB DRAG MODE
keyboard entries and program responses for the IND is invalid. This parameter is used by the program to
example case. The program will be executed from an in- decide whether the drag calculations should be perform-
teractive graphics terminal. ed using a planing hull method or Taylor Standard
Upon initiation of program execution, a title block is Series data. Valid values for this indicator are PLAN-
printed that contains the program name, version ING and TAYLOR. A planing hull drag calculation is
number, and release data of the version. The command requested by input to the current model of the value
prompt, CMD,E> indicates that the program is ready PLANING to the parameter HB DRAG MODE IND.
to accept input commands. The “E” in the command Execution of the hullborne hydrodynamic program is
prompt informs the user that all input and output data then requested. The result of program execution is the
will be in the English Unit System. An “M” would in- SPEED versus DRAGplot shown in Figure 6, which is con-
dicate the Metric Unit System. structed at the terminal within seconds following input
In this example case, the first command issued by the of the program call command HB HYDRO MODULE.
user is LIST COMMANDS. This will cause all general The graphics display remains on the terminal screen
executive commands to be printed, as shown. The next until erased by the user.
command issued by the user, SHIPS, queries the data The last command, EXIT, terminates program execu-
bank for the names assigned to complete current models tion. The current model, which is the ship MODEL 900
that have been stored. Eight ships are shown to exist in as updated by execution of the hullborne hydrodynamic
the data bank. program, is lost because no attempt was made to store
For the example exercise, the ship MODEL 900 will permanently the current model in the data bank.
be transferred from the data bank to the current model.
PROGRAM
USE:CASEHISTORIES
The USE, MODEL 900 command performs this task. A
single portion of the current model data will be examin- The HANDE computer-aided-design tool has been
ed. The ship length between perpendiculars is output by utilized by the David W. Taylor Naval Ship Research
simply entering the name of the parameter, LBP. The and Development Center (DTNSRDC), the Naval Sea
result, in feet, is shown. Metric output data will be pro- Systems Command (NAVSEA), and private firms con-
duced in the remainder of this example case. The cerned with hydrofoil ship design. It has been used for
METRIC UNITS command performs this task. The verification and modification of existing designs, for
length between perpendiculars, in meters, is then output research studies, and for rapid-response studies. For
in response to the LBP command. each of these tasks, the availability of the HANDE tool
The hullborne hydrodynamics computational pro- has enabled the Hydrofoil-Ship Engineering Communi-
gram will be executed to generate SPEEDversus DRAG ty to produce results more quickly and economically
data. The desired output is graphics display no. 1, which than would otherwise be possible. The many users and
gives a plot of SPEED versus DRAGfor three ships weights uses of the program, in conjunction with the increased
(full load weight, full load weight minus half fuel load, engineering productivity it affords, have validated the
and full load weight minus full fuel load). This display is basic concept of the HANDE tool.
126 Naval Engineers Journal, April 1981
KINGIDEVINE HANDE/COMPUTER-AIDED DESIGN APPROACH

Design Verification and Variation design, whereas in the HANDE design a conventional
foil system was used. The dramatic difference in foil
One of the most important uses of HANDE has been system weight resulted in a difference in the amount of
in the realm of verification or variation of manually- fuel that could be carried, which ultimately led to a
performed hydrofoil ship designs. In the area of design spectacular difference in foilborne range. Had HANDE
verification, HANDE has been quick to identify errors, not been available to generate economically the com-
areas that require additional consideration or develop- parable conventional design, the full impact of the in-
ment, and innovations and their impact. In the area of novation might not have been realized.
design variation, HANDE has allowed a rapid altera- Although not used in the HYD-2 verification effort,
tion or design data to reflect varied mission re- HANDE does have the capability to incorporate design
quirements or changing concepts. innovations in the design-generation process. This
Two series of hydrofoil ship designs illustrate the use capability involves substitution of user-derived data for
of HANDE for design verification and variation. The program-calculated data. Use of this capability is
first series was produced during the Advanced Naval discussed in a later section of this paper.
Vehicle Concept Evaluation (ANVCE) Study, and con- The Patrol Hydrofoil Missileship (PHM) design has
sisted of three designs. These designs were known as the been subjected to considerable amount of study regar-
HOC, HYD-7, and HYD-2. ding potential improvements. Proposed modifications
The HOC hydrofoil ship design has a full-load weight include substitution of propeller propulsion for the
of approximately 1,400 metric tons and has a multi- waterjet propulsion systems, use of diesel-driven
mission payload. TABLE 2 shows a comparison of generators for ship service power instead of gas turbine-
design data between the manual design and the HANDE driven units, and foil system modifications. These pro-
design. Differences are on the order of a few percent. posals were studied in the traditional engineering en-
The excellent correlation of design data resulted in in- vironment. HANDE was employed at a later date to
creased confidence in the HOC design. verify results of the variation studies. As expected,
Subsequently, some design requirements were altered. engineering analysts, through use of HANDE, were able
A new, manually-produced design effort would have to check results of the original studies in a fraction of
taken months to complete. HANDE was alternatively the time required by the original studies. And in one
used by the ANVCE designers to perform the design case that was studied in great detail, a design variation
variations. Estimated savings in engineering labor of PHM was found to have considerable conservatism
through use of HANDE instead of the manual effort that was not apparent in the original study.
was ninety percent.
Research Studies
The HYD-7 design verification effort by means of
HANDE highlighted some speculative characteristics of HANDE has been found to be invaluable as a
the design. HYD-7 has a maximum speed in the research tool. It has been used to perform parametric
supercavitating-foil range and thus required the use of studies whose scope could not have been achieved
exploratory, variable-geometry foil systems. HANDE feasibly by traditional means. It also has been used to
does not have the capability to analyze directly variable- perform economically detailed studies of key design
geometry foil systems and attempts to reproduce this issues.
design with HANDE failed. The failure emphasized the Two major parametric research efforts have been
speculative nature of the technology incorporated into done at DTNSRDC. One was called “Balancing
the foil system design. Mission Requirements a n d Hydrofoil Design
The verification effort of the 2,400 ton HYD-2 design Characteristics,” and the other was called “Hydrofoil
revealed the full impact of a technical innovation. As Operational Performance Envelope Extensions.”
TABLE 3 shows, the design had a very low foil system Results of the efforts were reported by CLARK, O’NEILL
weight compared with that calculated by HANDE. An and WIGHT[8], and by HAWKINS, KINGand MEYER [9]
innovative foil system design was used in the manual respectively.

TABLE 2

MANUAL VERSUS HANDE RESULTS FOR HOC

MANUAL RESULT HANDE RESULT MANUAL i HANDE

Length between perpendiculars (feet) 200 200 1 .OO
Military payload (long tons) 120 I20 1.oo
Lightship weight (long tons) 931 910 I .03
Loads (long tons) 504 530 0.95
Fuel weight (long tons) 42 1 446 0.94
Full load weight (long tons) 1,441 1,440 I .oo
Dynamic lift (long tons) 1,318 1,349 0.98
Foilborne design speed (knots) classified classified 1.oo
Foilborne power required (horsepower) classified classified 1.oo
Foilborne range (nautical miles) classified classified 1 .00

Naval Engineers Journal, April 1981 127

HANDE/COMPUTER-AIDED DESIGN APPROACH KING/DEVINE

TABLE 3

MANUAL VERSUS HANDE RESULTS FOR HYD-2

MANUAL
RESULT HANDE RESULT MANUAL i HANDE
Length between perpendiculars (feet) 320 320 1 .oo
Military payload (long tons) 279 279 1 .oo
Foil system weight (long tons) 233 537 0.43
Lightship weight (long tons) 1,491 1,919 0.78
Loads (long tons) 87 1 446 1.95
Fuel weight long tons) 653 23 1 2.83
Full load weight (long tons) 2,362 2,365 1.oo
Dynamic lift (long tons) 2,235 2,235 1.oo
Foilborne design speed (knots) classified classified 1.06
Foilborne power required (horsepower9 classified classified 0.88
Foilborne range (nautical miles) classified classified 3.05

The first research effort involved extensive develop- More recently, the benefits of foil system retraction
ment of four baseline ships in the usual interactive have been debated. Two similar ship designs were
mode. Then forty-eight variants were developed in the prepared quickly through use of HANDE. The sub-
less expensive batch mode. Quantitative relationships systems of these two ship designs were held in common
between mission requirements (speed, range, military as much as possible, but one had retractable foil systems
payload) and hydrofoil-ship design characteristics whereas the other had fixed foil systems. The impact of
(weight, power, size) were established t o provide an the fixed foil system on ship performance, size, and cost
awareness of the impact of Top Level Requirements. In was subsequently assessed by comparison of the design
contrast t o usual methods of performing parametric data for the two ships.
studies, HANDE provided a means for achieving a level Before the advent of HANDE, questions such as the
of integrated analysis and design that would otherwise impact of extended hullborne range or fixed-foil
not have been possible. systems could only be the subjects of speculation, or ex-
The second study demonstrated HANDE’s flexibility plored at great cost. HANDE has increased hydrofoil
in adaptation to concepts for which it was not specifical- ship designer productivity such that rational discus-
ly designed. The study evaluated the addition of large, sions, based upon inexpensively produced and rapidly
submerged buoyancy/fuel tanks to hydrofoil ships. This developed design data can take place.
concept provided a challenge t o the designer because Use of HANDE in the foil system retraction study
HANDE was not designed to accommodate such a produced an additional benefit. Design team par-
hybrid hydrofoil ship. The designers quickly discovered ticipants were able to function remotely from each
that although the design synthesis process could not be other. Technical interchange between members of the
employed directly, many of the HANDE computational design team occurred by telephone. Transfer of design
programs could be used in conjunction with manual data occured via HANDE and computer.
calculations t o calculate design data efficiently.
Seventy-two hybrid designs plus several conventional Rapid Response Studies
designs were evaluated. Based upon study results, the
NAVY plans t o demonstrate the hybrid concept full-scale HANDE has been unique as an engineering tool for
on the USS Highpoint (PCH-1) Hydrofoil Ship. developing rapid answers to questions that need im-
mediate answers. All too often designers are faced with
The suitability of HANDE as a research tool has been “what if” questions that require a response in a matter
demonstrated not only by parametric studies, but also
of hours or days. The amount of effort that can be put
by studies for resolution of design issues. This has been
into developing answers to such questions is obviously
demonstrated by a series of “medium size” hydrofoil
limited. HANDE provides a means by which tremen-
ship designs that were designed through use of HANDE
dous technological prowess can be quickly brought into
to an Outline NATO Operational Objective. The NATO the development of answers to such questions.
Corvette Hydrofoil (NCH) was the first of this series.
Two rapid response studies were performed at
Following completion of the NCH design, several ad- DTNSRDC which illustrate HANDE’s value as a rapid-
ditional configurations were derived in a short period of response tool. The Naval Material Command (NAV-
time. A key design issue regarding this series, that of ex- MAT) needed to ascertain the design characteristics of a
tended hullborne range, was subsequently raised. Via medium-size, multi-mission hydrofoil ship. NAVSEA
HANDE, designers were able to design rapidly a similar needed similar data regarding a large hydrofoil ship for
ship that was specifically configured for long hullborne NATO. The Hydrofoil Office at DTNSRDC was able,
range. The impact of the range requirement was subse- via HANDE, to respond to both concerns with com-
quently examined by comparison of the extended-range prehensive hull, structures, resistance, propulsion,
ship design with the earlier designs of the series. hydrostatics, weights, and cost data in less than a week.
128 Naval Engineers Journal, April 1981
KINWDEVINE HANDE/COMPUTER-AIDED DESIGN APPROACH

Without HANDE, this task would have been virtually Options,” Paper No. 78-750, AIAAISNAME Advanced
impossible. Marine Vehicles Conference, April 1978.

FUTURE
DEVELOPMENT
AND APPLICATION
APPENDIX
Since its initial implementation in 1977, HANDE has COMPUTATIONAL PROGRAM CAPABILITIES
undergone continuous enhancement. Interactive
graphics, the option of Metric or English input/output, INITIALIZATION - This program is normally the first program to
expanded hydrostatic analysis capability, new hullborne be exercised after assembling a new ship in the current model. Data is
resistance algorithms, and other smaller features have thoroughly checked for completeness and if no fatal errors exist
within the data, a mini-design synthesis process is initiated that con-
been incorporated into HANDE. Additional develop- tains geometric, hydrodynamic, propulsion, performance, and weight
ment, such as the ability to use controllable-pitch pro- calculation capability. Simple empirical methods are used throughout.
pellers and additional propulsion system options, will be The calculation sequence used by this program is as follows:
performed in the near future. Without continual
I . Input data are checked.
maintenance, the technology in HANDE, like an 2.Ship weight is estimated.
engineer without continuing educational experiences, 3.Hull is resized, if requested.
would become obsolete. Continued technological 4.Auxiliary and electrical systems are sized.
growth will allow HANDE to explore new horizons in 5.Foilborne and hullborne ship drag forces are calculated.
naval hydrofoil ship design. 6 .Foilborne and hullborne propulsion systems are sized.
7.Ship range or fuel weight are calculated.
Derivatives of HANDE have already been proposed 8.Ship weight is recalculated.
to cover the realms of Destroyer, Submarine, and Plan- 9.If the ship weight calculated in Step 8 does not approximately
ing Craft Design. Judging from the success of HANDE equal the weight as previously calculated, the mini-synthesis cycle
as a powerful engineering tool for hydrofoil ship design, is repeated from Step 3 until weight convergence occurs.
the development of additional tools of the HANDE type HULL GEOMETRY - The hull geometry program defines
for other ship types seems prudent. bulkhead, girder, and deck locations, and also defines superstructure
and hull geometry. Hull offsets in the current model are scaled and
REFERENCES warped t o define a new hullform that meets requested physical charac-
teristics. This program includes portions of the NAVY program
[I] Miller, R.T., “ A Ship Design Process”, SNAME “Hullform Derived from Parent.”
quarterly, Marine Technology (October 1965).
[2] Brennan, A.J., J.D. Burroughs, W.C. Hurt, W. Wichert HULL STRUCTURE - This module calculates scantling data for the
ship elements defined in the current model. The calculations are based
and D. Wacker, Hydrofoil Analysis and Design Program upon pressure loading data which are either calculated by the program
(HANDE) Phase 0 Final Report. Seattle, Wash.: The Boe- or input by the designer. Scantlings are determined at three
ing Company, D221-51302-1, June 1973. longitudinal locations for the hull bottom, hull sides, and weather
[3] A.J. Brennan, J.D. Burroughs, N.R. Knutsen, K.E. deck. Additional scantling data are calculated for lower decks,
Meldahl, D.E. Strain and D. Wacker, Application of the bulkheads, frames, girders, beams, and stiffeners.
Hydrofoil Analysis and Design (HANDE) Program -
Volume 0. Seattle, Wash.: The Boeing Company, FOIL/STRUT GEOMETRY - The foil/strut geometry program
D321-5 1321-5, September 1977. sizes foils, struts, and pods in accordance with the defined hull size
and with the foil system type and geometric data provided. Single T,
[4] A.J. Brennen, J.D. Burroughs, N.R. Knutsen, K.E.
double T , ?r or three-strut configurations may be used for the aft and
Meldahl, D.E. Strain and D. Wacker, Hydrofoil Analysis
forward foil systems. Longitudinal locations of struts are calculated
and Design Program (HANDE) Users Manual - Volume from a foil loading ratio specified by the designer.
1. Seattle, Wash.: The Boeing Company, D321-51312-1,
July 1976. HYDRODYNAMICS - This module uses hull and foil system data
[5] A.J. Brennan, J.D. Burroughs, N.R. Knutsen, K.E. to calculate foilborne drag and takeoff drag.
Meldahl, D.E. Strain and D. Wacker, Hydrofoil Analysis
and Design Program (HANDE) Theory Manual - FOILBORNE PROPULSION - This propulsion module performs
Volume II. Seattle, Wash.: The Boeing Company, sizing calculations for either a foilborne-waterjet or foilborne-
D321-51321-2, July 1976. propeller propulsion system. The waterjet-propulsion system section
of this program calculates engine power requirements, water-duct
[6] A.J. Brennan, J.D. Burroughs, N.R. Knutsen, K.E.
losses, pump size, and operating data based upon given drag, duct,
Meldahl, D.E. Strain and D. Wacker, Hydrofoil Analysis and pump type data. The propeller-propulsion system section
and Design Program (HANDE) Maintenance Manual - calculates engine power requirements, z-drive transmission
Volume I l l . Seattle, Wash.: The Boeing Company, parameters, propeller size, and propeller operating data based upon
D321-5 1312-3, September 1977. given drag, gearbox, and propeller characteristic data.
[7] A.J. Brennan, J.D. Burroughs, N.R. Knutsen, K.E.
Meldahl, D.E. Strain and D. Wacker, Hydrofoil Analysis HULLBORNE HYDRODYNAMICS - The hullborne
and Design Program (HANDE Data Bank Maintenance hydrodynamics program calculates ship drag data during hullborne
Manual - Volume I V , Seattle, Wash.: The Boeing Com- operation. Either planing hull or Taylor Standard Series drag-type
pany, D321-51321-4, September 1977. calculations may be performed.
[8] Clark, D.J., W.C. O’Neill, and D.C. Wight, “Balancing HULLBORNE PROPULSION - The hullborne propulsion program
M i s s i o n R e q u i r e m e n t s a n d H y d r o f o i l Design calculations parallel those of the foilborne propulsion program except
Characteristics,” Paper No. 78-725, AIAA/SNAME Ad- that all data for the hullborne propulsion system.
vanced Marine Vehicles Conference, April 1978.
[9] Hawkins, S., J . King, and J . Meyer. “Hydrofoil Opera- FOIL/STRUT STRUCTURE - The foil/strut structure program
tional Performance Enhancement Using Hybrid Design calculates scantlings of the primary load-carrying structure of the foils

Naval Engineers Journal, April 1981 129

HANDE/COMPUTER-AIDED DESIGN APPROACH KING/DEVINE

and struts. The calculations are based upon geometric data and upon HYDROSTATICS - The hydrostatics program determines the
loading conditions derived from hydrodynamic and inertial forces hydrostatic characteristics of a hydrofoil ship design. Data are
developed during foilborne operation. Loads include unsymmetrical calculated for hydrostatic properties of form, floodable length, intact
foil loading, hydrodynamic lift distribution, and incremental lift, stability, damaged stability, and maximum vertical center of gravity
drag, and side loads associated with maneuvers and operation in a sea positions allowed by NAVSEC Design Data Sheet DDS 079-1 criteria.
state. T h i s p r o g r a m i n c l u d e s p o r t i o n s o f t h e N A V Y“ S h i p Hull
Characteristics Program.”
FUEL/RANGE - Range performance is calculated by this program CONTROL SYSTEMS - This program allows the designer to obtain
in either of two ways. The weight of fuel required to achieve a quantitative information regarding the dynamic stability and con-
specified foilborne range is calculated or the range which may be trollability of the foilborne ship in a sea state. A set of stability boun-
achieved by a given ship is calculated. The calculation mode is daries based upon foil system geometry is calculated t o determine
specified by the designer. Fuel requirements for auxiliary and electric whether the boundaries are violated when the ship is exposed to
plants are also considered. several sea conditions.

WEIGHT - The weight program calculates a detailed weight COST -The cost program estimates hydrofoil-ship costs for the pur-
pose of design “trade-offs” and comparative evaluations. Both unit
breakdown for the ship. The weight statement follows the Navy Ship
production costs and life-cycle costs are addressed. Simple empirical
Work Breakdown Structure (SWBS).
relationships based primarily upon the NAVY SWBS are used to
estimate unit costs. Life-cycle costs are estimated utilizing a variety of
PERFORMANCE - The performance program calculates the per- data.
formance characteristics of ship designs that have been generated via
the design synthesis process. Whereas design-synthesis performance GEOMETRY DISPLAY - The geometry display module produces
calculations assume calm water and a clean ship, the performance pro- plots of ship geometry. Hull lines, bulkheads, decks, foil systems, and
gram considers fouling effects of marine organisms, degradation of superstructure can be assessed quickly and easily for correctness by
machinery with time, and sea state operation. the designer.

BOEING JETFOIL LAUNCHED FOR BELGIUM

THE FIRST OF TWOBOEINGJETFOIL HYDROFOILS for

Regie voor Maritime Transport (RMT), the state-owned
the name “Sealink, RMT will provide up to six trips a
day on the sixty-two nautical mile route when the se-
ferry company of Belgium, was launched by JETFOIL cond JETFOIL, Princesse Stafanie, is added to the ser-
manufacturer Boeing Marine Systems in Seattle, Wash., vice this summer. Direct train connections will enable
on 16 February 1981. travellers to make the trip between Ostend and London
in just three-and-one-half hours at approximately $43
one-way fare.
The 316-seat Hydrofoil, Princesse Clementine, will
enter commercial service on 3 1 May 1981 between RMT presently operates regular car-carrying vessels
Dover, England and Ostend, Belgium. Operating under between Ostend and Dover/Folkestone.

130 Naval Engineers Journal, April 1981