Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

SNAME Transactions, Vol. 95, 1987, pp.

29-44

DRYDOCK: An Interactive Computer Program for


Predicting Dry Dock Block Reactions
Ivan Jiang, 1 Associate Member, Kim Grubbs, 1 Visitor, Ross Haith, 2 Visitor, and
Vincent Santomartino, 1 Member

The DRYDOCK computer program was developed in order to overcome the limitations inherent in
existing dry dock analysis methods. Application of existing methods of analysis can be cumbersome
and extremely time-consuming, and have theoretical limitations created by practical considerations
of time and effort. The DRYDOCK program represents an important advance in the ongoing
evaluation of docking analysis methods. It is extremely user-friendly, and the ease of application,
flexibility, speed, and accuracy of the program makes it an attractive alternative to traditional
methods of analysis.

Introduction docking schedules, and loss of the ship to active duty until
repairs can be made. The problem is exacerbated by the
THE DRYDOCK computer program was developed un-
increasing complexity of hull forms and the proliferation
der the auspices of the Naval Sea Systems Command (NAV-
of tmusually shaped appendages such as sonar domes and
SEA) to overqome the limitations of existing dry dock
bulbous bows which make the application of traditional
analysis methods. DRYDOCK provides a quick, reliable forms of dry dock analysis more difficult and inefficient
method of estimating block loads when docking a ship in
than ever before.
either a graving or floating dry dock. Because of the po-
The DRYDOCK computer program was developed to
tentially disastrous consequences of docking block failure
overcome the limitations inherent in existing methods of
and the increasing difficulty in providing reliable docking
analysis and provide a reliable, efficient tool to be used in
analyses for large, modern ships, DRYDOCK was devel-
both graving and floating docking analyses. Designed so
oped as a tool for the design and optimization of efficient
that user-input data are kept to a minimum, the DRY-
docking systems. The DRYDOCK program is an important
DOCK data files for a particular ship, block arrangement,
advance in the ongoing evolution of docking analysis meth-
and dock are relatively simple to prepare. Once an initial
ods. data base has been prepared for a particular docking prob-
Historically, the development of tools for ship docking
lem, data base changes can be readily accomplished with
analysis has lagged far behind the development of tools
a minimum expenditure of time and effort. Thus, a docking
for other phases of ship design. While advanced methods
analysis can be quickly and easily performed at an early
of analysis are used in ship hull design, the problem of
stage of the design process, and potential problems can
drydocking a ship is typically addressed only after signif-
be identified early in the design.
icant ship parameters such as hull form, structure, and
weight distribution have been selected. This situation is
primarily due to the extreme complexity and time-con-
Previous studies
suming nature of traditional dry docking analyses. Current
methods of analysis are simply too cumbersome to be re- Early in the evolution of drydocking analysis methods,
peated and checked at each phase of the design spiral; the loading of docking blocks was assumed to be uniform
therefore, the docking analysis is perceived as a secondary over the range of blocks. The load on a particular block
problem to be addressed at the end of the ship design was determined by multiplying the calculated pressure by
process. The consequences of a poor drydocking analysis, the block/hull contact area. Unfortunately, this method
however, are potentially catastrophic. Docking block fail- of analysis is adequate only for relatively stiff, uniformly
ure while a ship is drydocked can cause not only loss of loaded hull forms supported over most of the length of
life, but also enormous economic losses. A docking failure the hull. For drydocked ships with large overhangs fore
can lead to extensive ship and dock damage, disruption of or aft, or ships with high local loads, the ship center of
gravity may not be vertically coincidental with the center
of pressure of the blocks, and results may not be reliable.
Designers & Planners, Inc., Arlington, Virginia. For this reason, a trapezoidal pressure distribution was
Surface Ship Structure Integrity Division, Naval Sea Systems assumed by Elgar [1] ~ in one of the earliest studies of a
Command, Washington, D.C. ship docked in a graving dock. This study, however, as-
Presented at the Annual Meeting, New York, N.Y., November
11-14, 1987, of THE SOCIETYOF NAVALARCHITECTSAND MA-
RINE ENGINEERS. 3Numbers in brackets designate References at end of paper.

29 "- J
sumes that the ship is perfectly rigid and does not bend such as local ship bottom flexibility and the effects of large
as a whole which may lead to unreliable results. Yeh and ship overhang are accounted for in the program.
Ruby [2] treated the ship as an isotropic, homogenous,
elastic beam supported by rows of elastic blocks resting
on a dock floor assumed to be infinitely rigid. The Ritz DRYDOCK assumptions
energy method was used in this study to estimate block
loads proportional to the ship deflection at each block The major assumptions made in the development of the
location. Palermo and Brock [3] used the assumptions of DRYDOCK program are as follows:
the Yeh-Ruby study to formulate a procedure for calcu- 1. The ship is modeled as a beam of varying cross section,
lating keelblock loads by treating the ship as a beam on and its weight is assumed to act at a number of discrete
an elastic foundation. points along the length of the ship. The longitudinal lo-
cations of these points, or nodes, are determined by the
Each of the studies mentioned neglects several factors
which are important when considering docking proce- block arrangement to be used in the analysis. At least one
docking block is located at each node.
dures for modern ships in a graving dock. Chief among
these is the assumption that the effect of ship overhang 2. The ship is supported by rows of discrete elastic sup-
ports, or nodes, which represent the docking blocks.
(either fore or aft) on the loading of the foremost or aftmost
blocks is negligible. For modern cruisers and destroyers 3. All materials are assumed to be linearly elastic. Keel-
which have large overhangs when docked, the combina- blocks are assemblies of different types of materials with
tion of ship and block flexibility results in an increase in different Young's moduli (E) and spring constants (K).
the extreme block reaction. The combined effects of ship 4. The dock floor is assumed to be perfectly rigid in a
and block flexibility, if ignored, can result in an unrealis- graving dry dock analysis. In a floating dry clock analysis,
tically low estimate of ship and block stresses. Also ne- the dock is modeled as a beam in the longitudinal direction
glected in these studies was the local flexibility of the ship and the pontoon is modeled as a plate with equivalent
plate thickness.
hull between transverse and longitudinal bulkheads, a fac-
tor which can lead to a variation in the load distribution 5. The ship/dock/block system is modeled as a rigid
over the docking blocks. Analysts in the design community beam supported by a row of elastic springs, resting on a
have attempted to use specialized structural computer rigid base. The flexibility of the ship bottom and the float-
programs (NASTRAN, STRUDL) to model the ship as a ing dock pontoon are taken into account in an approximate
manner.
flexible beam, but have found that the enormous effort
required to accurately model the ship bottom flexibility As stated above, the ship and dock are assumed to be
and block stiffnesses makes this approach undesirable. beams with nonuniform cross section. They are divided
When a ship is to be docked in a floating dry dock, the into a number of short beam elements, with the endpoints
problem is further complicated. The assumption of an in- of each element defined by nodes located along the length
finitely rigid dock floor is no longer valid, and additional of the ship/dock at every point where there is at least
steps must be taken to model the flexibility of the dock one docking block. The forward and aft ends of the ship /
itself and account for the forces which act upon it. Matrix dock are always the first and last nodal points.
formulation and computer generated solutions in the anal- To illustrate the manner in which equivalent blocks are
ysis of floating dry dock problems were first developed by utilized in the DRYDOCK analysis, consider a typical
Vaughan [4]. In this study, the ship was modeled as an transverse section of the ship/dock. At this section, the
elastic beam of varying cross section with its weight as- ship is supported by one centerline block and two (or more)
sumed to act at a number of discrete points. Two rows of side blocks. The three blocks can be considered to be
elastic supports arranged symmetrically on either side of elastic supports because of the assumption of all linearly
a center row of supports were used to model groups of elastic materials. Since the ship's bottom and dock's pon-
docking blocks; the dock bottom was modeled as a grillage toon are assumed to be rigid, the three blocks in the trans-
simply supported along its longitudinal edges and free verse row can be considered to be elastic springs in
parallel, as shown in Fig. l(a).
along its transverse edges. One limitation of this method
is that block removals must be symmetrical and the anal- With these assumptions, the three blocks can be com-
ysis is restricted to three rows of blocks; local deflection bined into one equivalent block with resultant stiffness of
k~ a, as shown in Fig. l(b). In this manner, a node is located
of the dock bottom, in the case of the floating dock, was
not considered. Potvin, Hartz, and Nickum [5] eliminated at each longitudinal point where there is at least one dock-
the lumping of transverse supporting members used by ing block, and equivalent springs representing the blocks
Vaughan, and addressed important factors such as varia- in the transverse row at that point are assembled. At the
tions of block stiffnesses, ballast and buoyancy variations,
random lack-of-fit conditions, and the crushing of docking
blocks. As with many of the previous studies on docking
problems, practical application of the theoretical aspects
of this study is limited by the enormous time and effort
which is required to perform calculations.
The DRYDOCK computer program described in this
paper was developed to overcome many of the limitations
inherent in the studies discussed earlier. The purpose of
the program is to provide a quick, reliable method to be
used to estimate block loads when docking a ship in a
//////////
graving or floating dry dock. The program combines many
of the theoretical aspects of earlier studies with the speed Keq= KI t-K2 t-K 3
and data reduction capabilities of the m o d e m computer (a) tb)
to produce a useful tool for the estimation of docking block
loads. Many important factors ignored in earlier studies Fig. 1 Block stiffness representation
30 DRYDOCK: An Interactive Computer Program
ends of the ship / dock, t h e r e m a y or m a y not b e a block. ~ p Sp3 "-~ Sp4
H o w e v e r , the ends can b e considered to b e s u p p o r t e d by
blocks of zero stiffness. I P~I
~sp| Es {p .GsAp Sp2~
Solution procedure
T h e solution process in the DRYDOCK p r o g r a m uses I- lp
the previously stated assumptions and input data con-
Fig. 2(e) Member deformation for a typical ship segment
cerning a particular ship, docking block a r r a n g e m e n t , and
dock to assemble a series of matrices based u p o n a gen-
e r a t e d set of linear, simultaneous equations. D e t a i l e d com-
m e n t s on the calculation of various e l e m e n t s of the
matrices a r e p r e s e n t e d later in this paper.
F o r a ship in the fully d o c k e d condition, the following
p•[ Fp3 "~ Fp 4
I P+!
equations are obtained based on the condition of equilib- ~Fp1 Fp2~
rittm:
[B] Y - - C - - F (1)
Fig. 2(b) Member forces for a typical ship segment
[D]Z= A + F (2)
(Note: Equation (2) is e l i m i n a t e d in the graving dock anal-
A typical b e a m s e g m e n t lying b e t w e e n the p t h and (p +
ysis d u e to the rigid dock assumption.) Notation:
1)th nodes is m o d e l e d as shown in Fig. 2. Notation for Fig.
Y= c o l u m n matrix of displacements at ship nodes 2:
Z = column matrix of displacements at dock nodes
F = c o l u m n matrix of the nodal reaction forces
E, = Young's modulus for ship
Ip = m o m e n t of inertia for s e g m e n t
= [K] (Y -- Z)
K = square matrix whose diagonal e l e m e n t s are equiv- Ap = shear area for s e g m e n t
G, = shear modulus for ship
alent stiffnesses of blocks
B = square matrix whose e l e m e n t s are stiffnesses of ship L, ----length of s e g m e n t
sp = deformations at e n d of s e g m e n t
beam elements
Fp = resultant forces at ends of s e g m e n t
D = square matrix whose e l e m e n t s are stiffnesses of
dock b e a m e l e m e n t s T h e stiffness coefficients matrix for the s e g m e n t is the set
C = c o l u m n matrix of l u m p e d ship weights at each node of relationships b e t w e e n the m e m b e r resultant forces and
A= c o l u m n matrix of resultant floating dock weights the m e m b e r deformations. T h e resultant stiffness matrix
and buoyancy forces. is shown in Table 1, where/3p is a t e r m which arises w h e n
shear deflections are considered, and is d e f i n e d as:
These equations r e p r e s e n t a set of simultaneous equa-
tions with twice the n u m b e r of nodes as unknowns. In EJp
general, t h e n u m b e r of unknowns will b e on the o r d e r of ~p - - Lp~G,Ap
hundreds. Fortunately, the coefficient matrix is a b a n d e d
sparse matrix with a population density of less than 5 per- This set of 16 n u m b e r s is called the m e m b e r stiffness ma-
cent. T h e stiffness and force matrices are assembled using trix k,, and gives the relation b e t w e e n force a n d displace-
the d a t a supplied by the user. m e n t for the p t h section. Its individual e l e m e n t s will b e
T h e t e c h n i q u e used in the p r o g r a m DRYDOCK for the r e f e r r e d to as ku(i,j = 1,2,3,4). W i t h this k matrix so de-
solution of simultaneous equations is called factorization. fined for the p t h m e m b e r , w e m a y now write F , = kpsp.
T h e subroutine p a c k a g e used to solve the equations has T h e resultant relation b e t w e e n the external n o d e force F
b e e n d e v e l o p e d b y Yale University [6]. T h e coefficient a n d the nodal d i s p l a c e m e n t r can be p r e s e n t e d as F = K
matrix is assumed to b e nonsymmetric. O n c e the matrices r, w h e r e K is the global stiffness matrix. F o r a t w o - s e g m e n t
are assembled, the sets of simultaneous equations are b e a m (Fig. 3), this matrix is shown in Table 2, w h e r e su-
solved and t h e displacements of the nodes are d e t e r m i n e d .
O n c e the translations are known, the deflections of the
equivalent elastic springs at each node are known. Mul- Table 1 Member stiffness matrix
tiplying the c o r r e s p o n d i n g stiffnesses and the vertical
translations gives the resultant block reactions at the FORCE CAUSEDBYA UNITOEFOR{WATIOOF:
N
RESULTANT
nodes. The reaction on each block in the transverse row
at a given node is t h e n found by distributing the resultant Sl s2 s3 s4
reaction in p r o p o r t i o n to the stiffness of each block in the 12Es|p -12Eslp 6Eslp 6EsIp
transverse row. FI Lp3(l*lZ~p) Lp3(l*12/~p) LpZ(I+12~) Lp2{1+12/~p)
Finally, a check on the calculations is m a d e b y s u m m i n g
the reactions at each block and c o m p a r i n g the sum with F2 -12Eslp IZEslp -6Eslp -6Eslp
the ship d i s p l a c e m e n t as obtained by integrating the ship Lp3(l+lZpp) Lp3(1+12~p) Lp2(l+tZ~p) Lp2(l+12#p)
weight distribution curve.
F3 6Eslp -6Eslp Eslp(4+12~p) 2Eslp(l-b~p)
Lp2(1+12~p) Lp2(1+12~p) Lp(l+12#~p) Lp(l+12~p)
The ship F4 6Eslp -bEslp ZEsIp(l-b~p) Eslp(4+iZ~p)
In DRYDOCK, the ship is v i e w e d as a nonuniform b e a m Lp2(I+lT-{3p) Lp(1+12~p) Lp(l+12~p)
d i v i d e d into a n u m b e r of discrete sections by nodal points. Lp2(].*lZ/?,p)
DRYDOCK: An Interactive Computer Program 31
r4 r5 r6 Maximum deflection for a single bottom
The single bottom is modeled as a stiffened plate with
two edges fixed and two edges simply supported. It is
assumed that the center girder is the only supporting
lr1 , r2 l r3 member in the longitudinal direction, and that the plate
is fixed by the transverse bulkheads.
Fig. 3 Degrees of freedom for a two-segment beam Based on the above assumptions and the input data,
various parameters for the stiffened plate are calculated
and deflection is derived by interpolating the curves given
perscripts 1 and 2 represent elements 1 and 2, respec- by Schade [7]. The orthotropic plate theory was extended
and applied to generate these curves.
tively. It should be noted that r~, r2, and r3 are vertical
deflections and r4, rs, and r6 are rotations.
Maximum deflection for a double bottom
Double bottom structure is modeled as a plate with two
S h i p b o t t o m flexibility edges fixed and two edges simply supported. The thickness
The block load distribution can vary greatly depending of the unstiffened plate is calculated by making its flexural
upon the local flexibility of the ship bottom structure. This rigidity equal to the moment of inertia of the box beam
variance suggests that the ship bottom can not be consid- formed by the inner and outer bottom plating. Based on
ered to be completely rigid. Considering the enormous these parameters, the deflection of the double bottom is
amount of data needed to adequately describe the ship calculated using the plate formulas presented by Roarke
bottom structure, the ship bottom flexibility is accounted and Young [8].
for in an approximate manner.
The basic approach to modeling the ship bottom is to
find the stiffness of the bottom and model it as an elastic P o n t o o n flexibility
spring. This elastic spring may be put in series with the The solution procedures described earlier are based on
equivalent elastic spring representing a transverse row of
the assumptions that the ship and dock are modeled as
docking blocks. The resultant spring may be taken as the
beams. Ship bottom flexibility is considered separately and
modified stiffness of the block at the node.
included in the block stiffness. Pontoon flexibility is esti-
To determine the spring constant of the ship bottom,
mated in a similar manner.
the bottom is assumed to be loaded with a uniformly dis-
The pontoon is modeled as a plate with four sides simply
tributed load. The maximum deflection is calculated at the
supported. The thickness of the plate is calculated by mak-
center of the bottom structure by assuming the bottom
ing its flexural rigidity equal to the moment of inertia of
structure to be simply supported along the sides and fixed
the box beam formed by the dock hull bottom plate, pon-
supported under the transverse bulkheads. A deflection
toon plate, and attached stiffeners. In order to simplify
pattern resembling a half cosine curve longitudinally and
the required dock input data, the moment of inertia is
a sine curve transversely is assumed based on these bound-
calculated from these plates, and then multiplied by 1.3
ary conditions. The deflection at the center of the bottom
to account for the inertia due to stiffeners. Formulas for
structure is then calculated based on the assumed deflec-
these calculations are given by Timoshenko and Woi-
tion pattern.
Once the block deflections are known, the load carried nowsley-Krieger [9] and by Pilkey and Chang [10].
The deflection of the simply supported plate with a
by each block is calculated. This calculation is done by
single line loading extending from x = a~ to x = a2 at a
finding the area of the ship bottom supported by each
distance b from the edge (Fig. 4) is given by
block and multiplying the area by the intensity of the
uniformly distributed load. ~, ~, n~rx mTrg
The total load carried by each block divided by the v = g,,,, sin sin
n=l m=l --~x Zy
deflection at the block is assumed to be the decoupled
stiffness of the bottom. This is assumed to be in series with where
the elastic spring representing the docking block, and the
modified stiffness of the docking block is calculated. Anm
For calculating the deflection at the center of bottom K,,, = D1r4 [(n2 / LxZ ) + (m2 / L 2)]~
structure, different approaches have been assumed for sin-
gle and double bottom subdivisions. These approaches will 4w mlrb f n~'al n~'a~
A,,,- n~rL~Sin~\c°s Lx cos L~ )
be described separately.

Table 2 Reduced structure stiffness matrix for two-segment beam

KII 1 k121 0 k131 k141 0


kEl I k221 + kll 2 klz2 kE31 kZ4 i + k132 k142
0 k~12 k222 O k232 kz42
K = . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
-- k311 K321 0 k331 k341 0
k411 k421 + K312 k3z 2 k431 k441 + k332 k342
0 k412 k422 0 k432 k442

32 DRYDOCK: An Interactive Computer Program


compressions of each layer of material in the block. The
/-L. /,_/ DRYDOCK method of calculation for knuckle reactions
works for either butted or spaced keelblocks.

Crushing loads
When block reactions are calculated in DRYDOCK,
/ these reactions are compared with the specified maximum
/ allowable load (crushing load). If a block reaction exceeds
/ the specified crushing load, then the load on that block is
/ fixed and the block is not allowed to carry any additional

?.,e / load. The DRYDOCK program will then recalculate the


block load distributions until no block reaction exceeds the
crushing load.
/- a2 ~/
Fig. 4 Loading on the dock plate element Program organization and application
The entire DRYDOCK package is written in FORTRAN
IV. The number of machine dependent subroutines has
It is usually sufficiently accurate to include only a few been kept to the barest minimum to ensure easy trans-
terms in the above equation. Linear superposition is used portability between various computer systems. The pack-
in the program to combine longitudinal rows of blocks in age has been developed and tested in the VAX 11/750
the docking process. and VAX 11/780 computer systems. Documentation for
Vertical deflections of the pontoon at a given point are the program includes the User's Manual for Program DR Y-
derived from the combination of block loads, dock weight, DOCK [12] and the Programmer's Manual for Program
and buoyancy forces. Nonuniform dock weight distribu- DR YDOCK [ 13]. Both the source code and documentation
tions and buoyancy loads are combined and assumed uni- for the program are available to the public, subject to the
formly distributed between blocks. The graving dock rules and restrictions of the Naval Sea Systems Command
analysis is first used to calculate the block loads. These for the release of such material.
loads are then distributed as line pressures along each row
of the blocks in the floating dry dock analysis. The pontoon Program DRYDOCK
deflections at the block locations are evaluated from the There are four separate programs which make up the
linear superposition of these surface and line pressures DRYDOCK package: (1) SHIPDT--Ship Data Preparation
acting on the equivalent plate. The resultant spring stiff- Program, (2) BLOKDT--Block Data Preparation Pro-
nesses (block, ship bottom, and pontoon) are then modified gram, (3) D O C K D T - - D o c k Data Preparation Program,
based on the relative deflection of the pontoon and the and (4) DRYDOCK--Main Program used for the analysis.
blocks. Theoretically, this procedure should be iterated The first three programs are interactive programs used to
because the block load distributions used in the pontoon prepare data files containing information concerning the
deflection calculation are derived from the graving dock. ship, block arrangement, and dock characteristics for the
However, because these deflections are due to the com- analysis. The fourth program uses the resulting data files
bined loadings, it was found that the change in resultant as input to carry out the DRYDOCK analysis. The sepa-
stiffness due to iterations is negligible. ration of input data into separate files allows existing files
for different ships, block arrangements, and dry docks to
be combined for a given DRYDOCK analysis.
Knuckle block reactions The three data preparation programs will be discussed
As a ship with trim aft is drydocked, the ship will first in greater detail in following sections. For the remainder
come into contact with the aftmost keelblock as the dock of this paper, "DRYDOCK" will refer to the fourth pro-
is pumped down. This block is generally known as the gram, which performs the docking analysis using the input
knuckle block. As more water is pumped from the dock, data files prepared by the user.
the decrease in ship buoyancy must be compensated by The DRYDOCK program is modular in construction,
an increase in knuckle reaction, so that at any instant the and consists of some 60 different subroutines, each of
sum of the knuckle reaction plus the remaining ship buoy- which performs a specific portion of the docking analysis.
ancy forces equals the ship's original floating displacement. Figure 5 is a flow chart detailing the logic used in the
Typically, the maximum knuckle reaction occurs when the DRYDOCK analysis and giving an indication of the func-
ship has lost approximately 50 percent of its original trim. tions of some of the more inportant subroutines. For a
The DRYDOCK program calculates the knuckle reac- more detailed analysis of the functions of each subroutine
tions produced when docking a ship with trim using the in the DRYDOCK program, refer to [13]. The modular
methods outlined in [11]. In the program, the initial ship construction of the program ensures that future modifi-
trim is successively reduced in ten equal increments, and cations to the program can be accomplished with a min-
the knuckle reaction is calculated at each step of the trim i_mum expenditure of time and effort.
reduction process. At each iteration, DRYDOCK deter- When the user begins execution of the DRYDOCK anal-
mines how many keelblocks are actually in contact with ysis program, h e / s h e is prompted for the names of the
the ship keel and determines the load, stress, and displace- three previously-created input files (ship, block, and dock).
ment of each block. Unlike the methods of [11] where the The DRYDOCK program then creates the following three
compression of each block is determined from a curve for additional files:
a typical docking block, DRYDOCK calculates the 1. The first file created by DRYDOCK is the output file.
compression of each loaded block by summing the In addition to an echo of all input data, the output file

DRYDOCK: An Interactive Computer Program 33


C~

START

HAIN PROGRAR
DRYNAIN v~s Jca11DBONJN . . . . . Calculate BonJean
-t data for shtp
ICall OlNINOJ . . . . . . . . . . . . . . I n i t i a l i z e Input/output
devices
ICall DBLKST . . . . . . . . . . . . . . . . . . . Calculate stiffness of each block
iCall DINERRJ. . . . . . . . . . . . . . I n i t i a l i z e variables tn
0-n ERROR Common Block
-< ICa111 ~ SUBROUTINE. . . . . . . Calculate knuckle
DKNBLK I DKNBLK block reactions
8 /~all ORDSHP/ . . . . . . . . . . . . . . Read Shlp Data Input File
~ ( ~ Call DWEIGT . . . . . . Calc. total weight
/Call ORDBLK/ . . . . . . . . . . . . . . Read Block Oata Input Ftle
l
/~a11 DRDDOK/. . . . . . . . . . . . . . Read Dock Data Input F11e

~ °
l | n i t t a l t z e variables in BLOCKS Common
JCall D[NBLO I . . . . . . . . . . . . . . Block; calc. max. allowable and
crushing loads for blocks
3
JCailOUNITIJ. . . . . . . . . . . . . . l n t t i a l l z e variables in UNIT Common Block
Blol

"u l
/Call DOUSHP//. . . . . . . . . . . . . . Prtnt ship tnput data to output f i l e i~ l Call DASCNDl. . . . . . Arrange blocks tn ascending
gl order w . r . t , ship LCF
3 / ~ a l l DOUBLy . . . . . . . . . . . . . . Print block Input data to output f t l e i Call D[NTPLJ...... Find shtp LCF and HTI
for midship draft
all DOUDOK//. . . . . . . . . . . . . . Print dock tnput data to output f i l e Calculate knuckle block
Jreacttons as shtp settles I
Ionto blocks and trim is j
JCall OCONVTI . . . . . . . . . . . . . . Convert input data to consistent system
of units for Internal calcs.
I reduced I

Fig. 5 DRYDOCK program flow chart


............ Calculate lengths of shtp beam elements
ICa'l -°°'*I -[ SUBROUTINE]. . . . . . . . Calc. stiffness of to be used In calculating shtp stiffness
DMOOST I shtp bottom subdtv|s|o,
stons

1 ............ Interpolate to f|nd wetght, mom. of |nertta,

I call DPTSTB1. . . . . . . . . Otvlde double bottom


tnto long. strips
V~
and shear area values fE::a:;:hshShste: bo;a:i:leae~rent

_ #SUBI OUTINE . . . . . . . . . . simultaneous eqns. for


DEOU grav|ng dock analysts
ICall DFAFT ~-. . . . . . . . talc. half-distances Calc. equivalent stiffness
0 between blocks I1 ............. of each transverse row of
30 blocks

.x.
I Calc. area of double bottom
Supported by each block I
I ............. Assemble stiffness matr|x
for ship beam model
p

ICalc. widths of divisions between I lEnA". . . . . . . . . . . Check block reactions to see t f maxtmum
long. bulkheadSsubdtvlstontn
a transverse allowable block load ts exceeded
N

Celt. max.
0 deflection
of bottom

NO
Calc. max. Check block reactions
"0 Ice11 DDBDEFi . . . . . . . . . . deflection ........ to see | f any exceed
of bottom max. a l l . a b l e load
g
3
ICalculate deflection I
of each block I
B /

ICall OPRINT~ . . . . . . . . . . . . . . Print block ~actlons, loads,


I Modify block sttffnesses I deflections to output f | l e
to refect stiffness of shtp
bottom
Y E S ~ ....... Create
IOPL~ plot f t l e

Flg. 5 (continued)
CO
U1
Sets up ltnear simultaneous equations for f l 0 a t i n g

EOU I was assembled in DGDEQUby SUBROUTINESHPSTF, i t


does not have to be redone here.

11-~. . . . . . . . . . . . . . Determine lengths of floating


<E~I drydock beam elements

.............. Determine dock deadweight loads to be loaded SUBROUT[ NE t . . . . . . . . . . Solve the set of simultaneous ltnear eqns.
onto dock beam element nodes DSOLVE using Yele Sparse Matrix Package (YSNP)
0 LL~. . . . . . . . . . . . . . Determine clock ballast loeds to be loaded
-n
-< onto dock beam element nodes
0 _] SUBROUTINE . . . . . . . . . . Part of YSMP. See
0 I--1 .I SUBROUTIN~ . . . . . . . . . . . Calc. bouyancy forces tO "J OORV YSMP documentation
C) 0v_J 7 ODBUOY | be loaded onto dock nodes
.x.
-I
J Call ORDER1 . . . . . . . . . Part of YSNP
g ~ ~
<.. or BonJeansJ
BONJEANS[ ~ N - - ] . . . . . . .
~DDBL~
Calc. buoyency
from BonJeens

......... Part of YSNP


1--~---- caic. beam ~9 1
element HYDROSTATICS J SUBROUTINE . . . . . . . . . Pert of YSI¢
o
3 st! ffness NPUT 7 NORV
"(3
C
g r

"o CO~DC~ ~ SUBROUTIN~ . . . . . Celc. buoyancy ......... Pert of YSMP


a ODHYOR J from hydrostatics
r

3 Call OOCHCK . . . .
Check weight and NN~ ......... Part of YSMP
hydro, data

NN~ ......... Part of YSMP


LISUBROUTINE I. . . . . . . . . . . . Calc. dock deflection due
-,DDKTRN J to transverse bending

.............. Calc. equivalent stiffness


of transvers rows of blocks

Fig. 5 (continued)
contains the results of the docking analysis, showing the i n p u t file. These options are chosen by t h e user d u r i n g
final deflections, loads, and stresses in each of the docking execution of the p r o g r a m and include t h e following:
blocks. In the case of a graving dock analysis, the o u t p u t 1. Use of default v a l u e s - - S H I P D T provides a n u m b e r
file will also contain the knuckle block reactions calculated of default values for Young's modulus and the shear mod-
as a ship with some initial t r i m first begins to settle onto ulus of steel, wood, and concrete in the British and S.I.
the docking blocks. systems of units. T h e user m a y use these values or p r o v i d e
2. T h e second file c r e a t e d by D R Y D O C K is the error his own as needed.
file. This file contains a listing of any errors d e t e c t e d by 2. Use of shear a r e a s - - S i n c e ships are relatively short
the D R Y D O C K p r o g r a m as the analysis is carried out. T h e c o m p a r e d to their depths, shear deflections can b e c o m e
D R Y D O C K p r o g r a m has a n u m b e r of built-in checks used important. T h e user has the option of ignoring or including
to d e t e c t inconsistencies in data and i n t e r m e d i a t e results. the effects of shear deflections w h e n S H I P D T is run.
Upon d e t e c t i o n of any data or i n t e r m e d i a t e result not 3. Ship b o t t o m f l e x i b i l i t y - - T h e user has the option of
following the e x p e c t e d pattern, an a p p r o p r i a t e error mes- including or ignoring the ship b o t t o m flexibility for single
sage is w r i t t e n to the output file, the error file, and the or double b o t t o m hulls.
screen of the user's terminal. A list of the possible errors 4. Calculation of buoyancy f o r c e s - - T h e buoyancy forces
which can b e d e t e c t e d by D R Y D O C K are included in [12]. acting on the hull m a y b e calculated e i t h e r from hydro-
If the D R Y D O C K analysis is p e r f o r m e d successfully, the static data or from Bonjean values. T h e user has the option
e r r o r file will b e empty. of inputting either hydrostatic data or ship offsets.
3. T h e third file c r e a t e d by DRYDOCK is the plot file, T h e d a t a n e e d e d to create a ship data file fall into the
which is c r e a t e d at the discretion of the user. This File following groups:
contains information n e e d e d to g e n e r a t e plots of the cal- 1. Identification data
culated block loads, deflections, or stresses for any longi- 2. G e n e r a l dimensions
tudinal row of blocks. 3. W e i g h t distribution
4. Ship structural data ( m o m e n t of inertia, shear area,
Program SHIPDT and ship b o t t o m structure data)
5. Hydrostatic data or ship offsets
S H I P D T is one of t h r e e data p r e p a r a t i o n p r o g r a m s writ- 6. Unit and material data
ten to aid the user in p r e p a r i n g input data files for use by A m o r e detailed explanation of the specific types of data
the D R Y D O C K analysis program. F e a t u r e s c o m m o n to n e e d e d as input for each group is contained in [12].
these t h r e e p r o g r a m s will b e first briefly discussed in this
section, and specific details r e g a r d i n g each p r o g r a m is t h e n
discussed. Program BLOKDT
T h e t h r e e d a t a p r e p a r a t i o n programs (ship, block, and
dock) r e m o v e the b u r d e n of p r e p a r i n g the data files in B L O K D T is the data p r e p a r a t i o n p r o g r a m used to aid
specific formats, thus eliminating r u n t i m e errors such as the user in t h e construction of the block data input file to
f o r m a t / d a t a mismatch. Each is an interactive p r o g r a m b e used with DRYDOCK. As with SHIPDT, it is an inter-
and is partially c o m m a n d d r i v e n in the sense that, while active, partially c o m m a n d driven p r o g r a m which allows
executing a c o m m a n d , it m a y p r o m p t the user for addi- the user to either create a n e w file from scratch or to
tional data. modify an existing block data file. T h e block data file cre-
In response to p r o m p t s from the programs, the user a t e d by B L O K D T will contain information on the location,
types in "'YES" or " N O " or a l p h a n u m e r i c data in any for- physical g e o m e t r y , and composition of each of t h e blocks
mat. T h e p r o g r a m s t h e n lay the data into data files in the in a particular docking block a r r a n g e m e n t plan.
necessary format. Additionally, each of the data p r e p a r a - B L O K D T is a particularly useful p r o g r a m d u e to the
tion p r o g r a m s contains a n u m b e r of checks to ensure that t r e m e n d o u s a m o u n t of data n e e d e d to adequately describe
data t y p e d in by the user are consistent with previously all of the blocks in a given docking a r r a n g e m e n t . In o r d e r
input data; e r r o r messages to alert the user of possible to c o m p l e t e l y describe a block for DRYDOCK, the d e p t h ,
p r o b l e m s with d a t a i n p u t are i n c o r p o r a t e d in the code of cross-sectional area, and material p r o p e r t i e s of each ver-
each p r o g r a m . tical layer of the block must be input. Since a typical block
T h e input data n e e d e d for the analysis are readily avail- a r r a n g e m e n t can contain h u n d r e d s of blocks, with each
able from various drawings, and can b e in either British, block c o m p o s e d of several different types of materials of
(S.I.) metric, or any o t h e r consistent system of units. T h e varying d e p t h s and areas, the a m o u n t of data n e e d e d can
p r o g r a m s take care of all conversions, constants, etc. ff b e c o m e enormous.
either the British or m e t r i c systems are adopted. F o r t u n a t e l y , B L O K D T has b e e n w r i t t e n in such a man-
Each of the t h r e e data p r e p a r a t i o n p r o g r a m s allows the n e r that this enormous a m o u n t of data can b e c r e a t e d with
user to modify an existing d a t a file by r e r u n n i n g the pro- relatively little input by the user. For example, a typical
gram. Additionally, existing data files m a y b e modified by block a r r a n g e m e n t m a y contain 70 equally spaced cen-
direct editting of the files. Like DRYDOCK, each of the terline blocks of similar composition; each of these blocks
t h r e e data p r e p a r a t i o n p r o g r a m s is m o d u l a r in construc- m a y b e c o m p o s e d of 6 layers of material stacked one u p o n
tion, increasing the ease with which modifications to the t h e other. Instead of inputting data for each block, the
p r o g r a m s can b e i m p l e m e n t e d . user n e e d only specify the longitudinal location of t h e first
S H I P D T is the data p r e p a r a t i o n p r o g r a m used to aid and last blocks in the row, the block spacing, a n d the
the user in p r e p a r i n g the ship data input file. It has b e e n composition of one of the blocks; the B L O K D T p r o g r a m
w r i t t e n in such a m a n n e r that all data t y p e d in by the user will t h e n insert each of the 70 blocks along t h e c e n t e r l i n e
in response to p r o g r a m p r o m p t s will b e layed into the ship at the p r o p e r location. Modifications to the composition
data input file in the correct location and format to b e of individual blocks in the row, or the addition of blocks
r e a d by the D R Y D O C K analysis program. at locations other than those already input m a y be accom-
A n u m b e r of options exist within S H I P D T as to the t y p e plished easily using various c o m m a n d s w r i t t e n into the
and a m o u n t of data which will b e layed into the ship data B L O K D T program.

DRYDOCK: An Interactive Computer Program 37


If the user does not wish to define the components of t~
I
a block, BLOKDT provides the option of inserting Navy
Standard Blocks at a given location. A Navy Standard Block / 't
in BLOKDT is defined as a block consisting of a four-foot- ,,,/
high layer of concrete sandwiched between six-inch layers
of wood. If the user desires, BLOKDT will insert whole
Navy Standard Blocks, stacked one on top of the other, a
between the dock floor and the ship hull at a particular
o
longitudinal location; any remaining space between the a

ship hull and the top of the stacked blocks will be filled D
o
with wood. If desired, a soft cap can be placed on the top o
[]
of any or all docking blocks. 8
At present, BLOKDT limits the number of docking i-B=
blocks in a given arrangement to 600. Each block may be D
. a =
composed of up to ten layers, and a maximum of five D

different types of materials can be used in the block com- xgr,


0

positions. Sonar pits in the dock floor and soft caps on the D

docking blocks may be input while running BLOKDT. l =

Program DOCKDT g 1

DOCKDT is the data preparation program used to pre- u


o
[] =
a
pare the dock data input ides for DRYDOCK. This pro- D

gram can prepare data files for either graving or floating a


a
a 0 o
dry docks, but the data needed to prepare a graving dock 0

data file are minimal and consist only of the major di- D
0
mensions of the dock. The data needed in preparing a
floating dry dock file are similar to those needed for the
ship and include information on the dock weight distri-
bution, ballasting plan, hydrostatic data, and the like. As
with SHIPDT and BLOKDT, DOCKDT has been written
to minimize the possibility of fatal errors occurring due to
format mistakes or inconsistencies in input data.
The structure and organization of DOCKDT are very
similar to that of SHIPDT, and will not be examined in
detail here. Like SHIPDT, shear deflections may be in- Fig. 6 Docking block arrangementfor typical destroyer
cluded or ignored, and the buoyancy forces on the dock
are calculated either from hydrostatic data or Bonjean
data.
of the ship at the time of docking was taken to be 6969
The major groups of data needed to create a floating metric tonnes (t), and the composition of a typical keel-
drydock data file are: block used in the test procedure is shown in Fig. 7. The
1. Identification data instrumentation and test procedure used in the measure-
2. General dimensions ment of loads in keelblocks nos. 1 and 3 are detailed in
3. Dock deadweight distribution [15].
4. Dock ballast data For comparison with the measured keelblock loads de-
5. Dock structural data (moment of inertia and shear termined in the BIW test, a DRYDOCK analysis for the
area data) destroyer was performed. Ship, Block, and Dock data input
6. Dock hydrostatic data or Bonjean data fries were prepared using the programs SHIPDT,
7. Unit and material data BLOKDT, and DOCKDT, respectively. These input files
Specific data needed for each of the above groups is de- were then used to execute the DRYDOCK analysis pro-
tailed in [12]. gram, and the predicted block loads were compared with
the Bath Iron Works results.
A comparison of the DRYDOCK predicted keelblock
loads with the Bath Iron Works measured keelblock loads
V a l i d a t i o n o f D R Y D O C K results shows that DRYDOCK can be used to accurately predict
To increase the level of confidence in the results gen- block loadings. The BIW test resulted in measured keel-
erated by the DRYDOCK program, the results of a DRY- block loads of 353 t on the aftmost block (Keelblock 1) and
DOCK analysis were compared with test data obtained 334 t on Keelblock 3; the loads predicted by DRYDOCK
~c,An the measurement of actual keelblock loads for a typ- for these two blocks were 361 t and 296 t for Keelblocks
ical destroyer with large fore and aft overhangs. The re- 1 and 3, respectively. It is of interest to note that BIW also
suits of this comparison are given in this section and show performed an analytical study using the structural com-
that DRYDOCK can be used with confidence. puter program STRUDL in an attempt to predict the keel-
The test case used for validation of the DRYDOCK pro- block loads produced in the docking procedure. The
gram involved docking a typical destroyer in a floating results of this analysis (314 t for Block 1 and 291 t for Block
dry dock. The measurement of keelblock loads was per- 3) seem to underpredict the actual block loads as measured
formed by Bath Iron Works (BIW) in 1984 at Bath's Port- in the test. Table 3 summarizes the results of the DRY-
land facility. The blocking arrangement used for this test DOCK analysis, the Bath Iron Works NASTRAN analysis,
is detailed in [14] and is shown in Fig. 6. The displacement and the BIW measured loads for Keelblocks 1 and 3. Figure

38 DRYDOCK: An Interactive Computer Program


12" SOFT CAP Table 3 Predicted and measured loads
YELLOW PINE

l oL OF SHIP Keel Block Loads (Metric Tonnes)

BIW STRUDL DRYDOCK BIW Measured


Model Analysis Loads

I
OAK
I
OAK
T Keel Block #I 314 361 353
Keel Block #3 291 296 334
CONCRETE! CONCRETE

I [.7.~
L I I IirJ
4"x4"x I / 4 " ANGLE,P&S sition or material properties of blocks, or the effects of
CONCRETE CONCRETE
SPD 13"-0 ° various ballasting schemes in floating dry dock analyses)
and to optimize the docking solution. Additionally, ship,
OAK OAK 1 / 4 " x 4"F B (TYP..L_ block, or dock data Files from different docking analyses
QAK QAK ~ TIE PC can be combined in any manner in order to find the best
ship / block / dock combination.
CONCRETE CONCRETE 3. Speed--As mentioned, input data files for a particular
4" ship, block arrangement, or dock are extremely easy to
OAK OAK create and require a minimum expenditure of time by the
OAK OAK
user. The DRYDOCK program performs internally many
CONCRETE CONCRETE 20" 2'-6"
of the calculations which have made previous methods of
analysis so time-consuming (calculation of block stiffnesses,
OAK OAK t ~ PONTOON insertion of blocks, etc.). The DRYDOCK analysis program
utilizes efficient, fast-running algorithms in the solution
. 2'.0" 6"
process.
4. Program modification--The modular construction
and extensive documentation of the DRYDOCK program
makes it extremely easy for a programmer to make future
Fig. 7 Typical keelblock modifications to the program in order to test varying as-
sumptions and methods of data reduction.

8 is a plot of the keelblock load distribution generated by 37O


the DRYDOCK analysis and the BIW NASTRAN analysis, 36O
..,,1--- KEEL BLOCK # 1
with points representing the BIW measured loads for 35O
Blocks 1 and 3 shown. As shown in Fig. 8, DRYDOCK 34O
33O P,,e--KE EL BLOCK ./¢3
tends to predict higher block loads at the fore and aft ends
320
of the row due to the large ship overhang at the ends. 310
"'- BIW NASTRAN PREDICTED LOADS
The results of this initial comparison of DRYDOCK loads 3OO
-- DRYDOCK PREDICTED LOADS
with measured loads are encouraging; while it is true that 290
28O • BIW TEST RESULTS
DRYDOCK seems to underpredict the load on Block 3 in 270
this case, it is very accurate in the prediction of the load 260' L~
on the aftmost keelblock, which is often the most critical 250'
block in terms of block failure. To increase confidence in 240, L~
DRYDOCK analysis, additional drydocking tests should be
z
230. L,
220,
performed and compared against DRYDOCK analyses. o
p- L,
210.
(.} ILI
200-
]90
i
Conclusions 180 0
170 I
The DRYDOCK program is superior to traditional meth- 160
I
ods of analysis in a number of important areas: 150
I. Data preparation--The DRYDOCK data preparation .
140
programs make it extremely easy to prepare input data g~ 13O
files for a particular ship, block arrangement, and dock for 120
110
use with the DRYDOCK analysis program. These data 1DO
t: ..I •
preparation programs are simple to use and ensure that 9O
L,
the user spends a minimum of time setting up the initial 8O
database for a particular docking problem. 70
2. Flexibility--The DRYDOCK program has been con- 6O

I
5O
structed to allow the user to examine a wide range of 40
solutions for a particular docking problem quickly and 3O
easily. Once the initial database for the ship/block/dock 20
system has been created, changes to this database are easily 10
performed. Thus, the user can perform a large number of 5 10 15 20 25 30 35 40 45 50
analyses to examine the effects of relatively subtle changes
KEEL BLOCK NUMBER
in the overall docking scheme (such as the removal or
relocation of key docking blocks, changes in the compo- Fig. 8 Keelblock load distribution for a typical destroyer

DRYDOCK: An Interactive Computer Program 39


5. Accuracy of results--DRYDOCK takes into account pated, however, that there may be instances where high
many factors which have been ignored in previous studies local loadings in the ship (such as tank loads, machinery
due to the difficulty and impracticality of applying ana- loads) may produce high localized loads in some docking
lytical tools to these factors. These include the effect of blocks. This problem should be investigated.
large ship overhangs, the modeling of the ship and dock.
flexibility, and the effect of shear deflections on block
loadings. Initial comparison of DRYDOCK results with Acknowledgments
measured block loads and other methods of analysis show
The authors wish to express their appreciation to Naval
that DRYDOCK can be used with a high level of confi-
Sea Systems Command for supporting the development
dence in the results.
of the DRYDOCK program and to Designers & Planners,
Inc. for their interest and encouragement. The authors
also wish to express their thanks to Mr. Yusuf Officiwala
Results and recommendations and Ms. Iris Yamamoto for their contributions in the de-
Initial comparison of DRYDOCK results with available velopment of the DRYDOCK program. Special thanks go
test data is encouraging; the loads predicted by the DRY- to Ms. Lorenna Hess for her help in editing this paper.
DOCK program compare favorably with measured loads
obtained by Bath Iron Works for a drydocking case for a
typical destroyer. Additionally, the loads predicted by
DRYDOCK seem to be more accurate than those obtained
using a NASTRAN model for the same docking case. In
References
order to increase the level of confidence in the DRYDOCK 1 Elgar, F., "The Distribution of Pressure Over the Bottom
program results, it is recommended that additional test of a Ship in Drydock and Over the Dock Blocks," T•ANS. SNAME,
Vol. 16, 1899.
data be obtained for a number of different ships and docks
2 Yeh, G. C. K. and Ruby, W. J., "A New Method for Com-
and compared with DRYDOCK analysis results. The dis- puting Keel Block Loads," TRANS.SNAME, Vol. 60, 1952.
tribution of the DRYDOCK package will be dependent 3 Palermo, P. M. and Brock, J. S., "Investigation of Pressure
upon further verification of the program results and will on Keel Blocks During Drydocking of USS MIDWAY(CVA41),"
be in accordance with NAVSEA's distribution procedures. David Taylor Model Basin Report No. 1003, April 1956.
There are a number of modifications which can be made 4 Vaughan, H., "Elastic Analysis of a Ship in a Floating
Dock," TRANS.SNAME, Vol. 108, 1966.
to the DRYDOCK program to increase its usefulness in 5 Potvin, A. B., Hartz, B. J., and Nickum, G. C., "Analysis of
future docking analyses. It is reeommended that modules Stresses in a Floating Drydock Due to a Docked Ship," TRANS.
be incorporated into the program in order to calculate and SNAME, Vol. 77, 1969.
predict the stability characteristics of a floating dock / ship 6 Eisenstat, S. C., Guisky, M. C., Schultz, M. H., and Sherman,
combination as a ship is docked in a floating dry dock A. H., "The Yale Sparse Matrix Package II. Non-Symmetric Prob-
facility. Similarly, calculations should be incorporated to lems," Yale University, Department of Computer Science, Tech-
nical Report No. 114, 1977.
examine seismic effects produced by earthquake activity 7 Schade, H. A., "Design Curves for Cross-Stiffened Plating
for ships docked in graving docks. Under Uniform Bending Load," TnANS. SNAME, Vol. 48, 1940.
Another area of study which should be investigated and 8 Roark, R. J. and Young, W. C., Formulas for Stress and
applied to the DRYDOCK program involves the accurate Strain, 5th ed., McGraw-Hill, New York, 1975.
determination of material properties for various types of 9 Timoshenko, S. and Woinowsky-Krieger, Theory o f Plates
and Shells, McGraw-Hill, New York, 1959.
wood used in the construction of docking blocks. The elas- 10 Pilkey,W. D. and Chang, P. Y., Modern Formulas for Stat-
tic modulus of various types of timber can vary greatly ics and Dynamics, McGraw-Hill, New York, 1978.
depending on a number of factors such as the age of the 11 Naval Ship's Technical Manual, Docking Instructions and
wood, whether the wood is wet or dry, and the effects of Routine Work in Drydock, NAVSEA89086-79-STN-000, Chapter
previous loadings on the proportional limit of the wood. 997, Dec. 15, 1977.
It is recommended that a series of docking analyses be 12 User's Manual for Program DRYDOCK, Designers and
Planners, Inc., Arlington, Va., 1987.
performed, using a wide range of modulus values, to de- 13 Programmer's Manual for Program DR YDOCK, Designers
termine the effects of varying elastic modulus values on and Planners, Arlington, Va., Inc., 1987.
the load-carrying capabilities of docking blocks. 14 Docking Plan; USS SCOTT DD-995, USS CONOLLY DD-
Another option which should be incorporated into fu- 979; Docking Position #2; Portland Facility; BIW Dwg. No. HE-
ture versions of the DRYDOCK program is the effect of 011184 Revision D, Bath Iron Works Corp., Bath, Me., Jan. ll,
high local loadings in the ship hull on the block loads. At 1984.
15 "Analysisand Evaluation of Instrumented Dry Dock Block
present, DRYDOCK uses the standard weight distribution Loading Test," Project No. 32-261-006, prepared by Giannotti
for a ship when performing the load analysis. It is antici- and Associates, Inc., Annapolis, Md., Oct. 6, 1984.

Discussion
Orhan Gurkan, Member, and George Kapsilis, Member and used trapezoidal weight distribution for naval com-
batants with large overhanging ends in drydocking cal-
The authors are to be commended for presenting a sig- culations, and it deserves the close attention of those
nificant contribution to one facet of drydocking operation. responsible for the structural design of ships and docks,
The phenomenal difference in the docking block loads and of those engaged in the drydocking operations.
experienced by the knuckle points and those elsewhere The paper makes it clear that the ship and dock scan-
along the ship's length, as illustrated by the DD-963 ex- tlings must also be designed to withstand the high docking
ample, brings forward several considerations: It indicates loads at the knuckle points. In regard to structural dry
the necessity of departing from the traditionally accepted dock design, the previous approach of basing the scantlings
40 DRYDOCK: An Interactive Computer Program
on a uniform load per foot must be reexamined: Stronger tude of the load distribution along the length of a ship
transverse girders or transverse frames with smaller spac- supported by a typical blocking system. As there are a
ings must be considered in way of the aft and forward number of uncertainties in the properties of the block
knuckle areas. materials, and the structure of the ship and the dock, the
The knuckle loads in the case of the DD-963 and other actual load in a given block may vary greatly from the
similar ships impose compressive stresses, perpendicular predicted value. Still, this program will be of great value
to grain, which are well above the proportional limit for in developing hull structural arrangements and details,
the oak timbers. These blocks, and those adjacent to them, particularly in ships with extreme overhangs. The prep-
must be built of either suitable synthetic materials or of aration of docking drawings and the evaluation of unusual
the traditional material of wood in end-grained small blocking arrangements necessitated by hull structural re-
squares bonded with epoxy or other appropriate adhesives pair will also benefit from the availability of this tool. The
(wood-epoxy composites). Douglas fir or yellow pine would authors are to be congratulated for their contribution to
seem to be the most suitable for this purpose since oak our understanding of a difficult problem.
does not agree with most adhesives. The rigidity of the
end-grained bonded timber can be overcome by the use Richard D. Hepburn, Member, and James K. Luchs, 4
of thicker, soft caps. We are of the opinion that experi- Visitor
menting with end-grained knuckle block timbers would [The views expressed herein are the opinions of the discussers
produce fruitful results in resolving the problem of ov- and not necessarily those of the Department of Defense or the
ercoming excessive loads and stresses at the knuckle Department of the Navy.]
points.
Finally, we at Rosenblatt have been involved in the Congratulations to the authors for their outstanding ef-
technical review of floating dry dock safety certification fort in the creation of this drydocking analysis tool. Older
reports for the past ten years and have recommended the ships had simpler hull forms, were fully supported along
deletion of the trapezoidal longitudinal weight distribution the keel, had low side blocks, and no large overhangs;
method prescribed in MIL-STD-1625B(SH). It is hoped therefore, simple analysis methods were adequate. Now,
that this paper will encourage Naval Sea Systems Com- with the advent of nuclear ships and surface ships with
mand to adopt the new approach for docking load analysis large overhangs and high side blocks, there is a require-
in their continuing efforts to improve the U.S. Navy Dry- ment for high confidence in the dry dock blocking systems.
docking Facilities Safety Certification Program. Even for some the older ships there has been a need
for a more accurate way of calculating loads on specific
David M. Maurer, Member dry dock blocks. The aft knuckle blocks on a battleship
had to be removed in order to release the stern tube
The degree to which the program simplifies the bottom bearings. It became critical to know the exact loading on
structure does not appear suited to configurations such as the next keelblock once the end blocks were removed. At
docking keelsons, stanchions, longitudinal bulkheads, that time no sufficient way was available to accurately
skegs, and rounded or " v e e " section shapes. The limita- calculate this. DRYDOCK appears to be the perfect an-
tions imposed by the manner in which the program models swer to the problem.
the flexibility of the bottom structure should be clearly There are few potential problems with DRYDOCK
stated. which need to be discussed. A large number of floating
The materials used to build up dry dock blocking systems dry docks in use today are extremely old and are largely
have been the subject of some recent study by NAVSEA. 3 made of wood. The timbers used in these dry docks have
Strength testing of" old" and " n e w " timbers, various grain been reloaded for years. Modeling the floors of these dry
orientations, configurations of layers, and moisture and docks as a single plate may be too simple an approximation
temperature conditions were used to determine the real- due to the high nonlinearities of wood. For steel floating
world strength properties of the woods commonly used in dry docks, the assumptions and calculations appear valid.
blocking systems. It was found that the properties of in- As the authors mentioned, the material properties of the
terest--fiber stress at proportional limit and modulus of wood used in drydocking blocks also vary considerably.
elasticity--vary considerably with time in service and the This is currently being examined in the seismic study at
conditions of use, and that the actual average values for M.I.T. and research ongoing at the University of Wash-
these properties are much less than expected based on ington.
published data. The variability within species is significant. The stiffness matrix method is an excellent approach in
The findings also indicate that timbers that have been determining deflection of the ship, block, and dock system.
stressed beyond their proportional limit but not visibly A similar method is being used to determine the horizontal
crushed have a significantly reduced modulus of elasticity block stiffnesses for the seismic loading study. These hor-
and in the next reuse will not carry their share of the load. izontal stiffness calculations could be incorporated into the
In light of the above, the default value for the elastic DRYDOCK program.
modulus of wood in the program may be worth reconsi- Conclusions about the accuracy of the DRYDOCK pro-
dering. Also, the assumption that the load on a block may gram based on only two experimental data points are ten-
be frozen when the block reaction exceeds the crushing uous. Measurements should be taken along the entire keel
load bears scrutiny. It may be more appropriate to consider line on ships in graving and floating docks. The limitation
such a block as completely unloaded. of ten layers for blocking configurations is too restrictive.
It would appear that the results given by the DRYDOCK For high buildups, eleven layers can be present. DRY-
program are indicative of the general shape and magni- DOCK's input programs are much better than manual data
entry but cannot be characterized as "extremely easy" to
use. A major revision of reference [ 11 ] (NSTM 997) that
3"Preliminary Results--Transverse Compression Studies of
Drydocking Blocks and Components," Naval Sea Systems Com- 4Lieutenant, USN; Department of Ocean Engineering, M.I.T.,
mand, Washington, D.C., Aug. 28, 1985. Cambridge, Mass.

DRYDOCK: An Interactive Computer Program 41


is currently in progress will significantly change the al- the Naval Architectural Branch has studied each ship's
lowed knuckle loading. These changes need to be incor- docking drawing prior to use to evaluate the possibility of
porated into DRYDOCK. reducing the number of blocks required. Our simple stud-
ies have found that many of the Navy's largest ships with
fiat bottoms, uniform loading, and more than a single row
John W. McGruer, Member of keelblocks can be docked safely with fewer blocks than
the standard docking plan indicates. These ships normally
The presentation of this paper and the DRYDOCK com- have six-foot spacing center-to-center for keelblocks. We
puter program is a major step in the development of tech- have achieved considerable cost saving and reduced dock
nology coupling computer aides and the safety and preparation time by utilizing eight-foot spacing of keel-
predictability of docking modern ships. Advances in naval blocks. Also, with the new spacing the underwater hull
architectural technology are becoming evident in modern can be painted in two dockings versus the normal three.
ship design; however, inherent with that design, little at- We have been restrained from considering cost saving
tention has been focused on the ship's capability to be blocking arrangements on some of the more complex hull
drydocked or to the concentrated loads imposed during forms because we did not have an accurate technique to
dry dock operations. The DRYDOCK computer program evaluate the best block / ship / dry dock combinations. With
can provide a useful method of highlighting, for the ship DRYDOCK now available, we are looking forward to ana-
designer, indications of high loads to supporting areas lyzing each ship we dock to optimize the docking block
where additional structural emphasis must be applied. The arrangement. We hope that we will be able to eliminate
computer program for the practical end users such as dock- some of the blocks in overdesigned areas and thus reduce
hag officers, docking engineers and doekmasters falls short cost and dry dock preparation time. In some cases DRY-
of the mark because of broad assumptions applied which DOCK analysis may find additional blocks are needed, but
are not compatible with conditions found during a docking in ten years we have docked only one class of ship that
evolution. required additional blocks to be added to an existing dock-
The assumption that modern ships can support a cal- ing drawing. This was due to the ships' having a very large
culation of uniform load distribution to determine the load stern overhang along with a heavy stern gate. Basically,
at each keelblock is feasible only on paper when loads are docking arrangements have been very conservative. This
calculated by the trapezoidal rule method. Considering was due in the past to the lack of an adequate technique
that this method has been applied for years using slide to obtain results with a high level of user confidence. As
rules and hand calculators, it is an inappropriate approach a result, many ships probably have overdesigned block
to apply similar assumptions to a sophisticated computer arrangements in some areas. DRYDOCK, we hope, is the
program. Simple superimposition of a modern ship (DD- tool to provide more cost-effective docking block arrange-
963 or CG-47) 20 station weight breakdown over the same ments without compromising safety.
ship's trapezoidal diagram will clearly indicate the areas For users to gain confidence in DRYDOCK, additional
of high loading and the lack of uniformity of load distri- comparisons of the program's results against actual ships'
bution. block loading data will be needed. I suggest that the au-
Assumptions regarding the elasticity of the supporting thors utilize reference [3], data on the USS Midway's ac-
keel blocking system are unfounded and at best extremely tual pressure on keelblocks, to compare results. The
difficult to predict. Recent studies by Naval Sea Systems benefit of these data is that many blocks throughout the
Command conducted at the University of Washington in ship's length were tested.
Seattle have disclosed that Young's modulus within indi- Norfolk Naval Shipyard is eager to obtain DRYDOCK.
vidual timber species cover a wide range. This factor plus It would be a privilege to assess for NAVSEA and for the
the wide variations in methodology across the continent authors the ease and speed of data preparation, and the
in erecting a keelblock buildup deters from the assumed flexibility of examining a wide range of solutions for a
linearity in the keel elastic assumption. particular ship.
The docking of ships and load distribution have for a
long time been treated with little concern; as we move Paul S. Crandall, Member
forward, however, there is a realization that in the past
the complexities of docking ships have been underesti- I wish to commend the authors for a very interesting
mated. There is therefore a desperate need for increased paper. The field measurements by BIW of block loadings
investigations and docking systems development as well of a Spruance Class destroyer in their huge AFDB floating
as a need for applied computer technology and instru- dry dock at Portland, Maine, showing good correlation of
mentation to establish existing conditions. The DRYDOCK the predicted loads versus tests, do not seem to have been
computer program is the breakthrough and will, it is done for the full length of vessel. I wonder if blocks 20,
hoped, serve as the springboard to the future. 35 and 50 had as good results?
My compliments to the authors for their interest in the I would ask the authors to comment on the effect of
subject of drydocking ships and for their contribution in solar heat in their program since we have seen ships lift
advancing specialized technology. clear off their blocks at bow and stern in the cool of night
after a very hot, sunny day. A paper of mine printed in
the August 1985Journal of Ship Production addresses this
Marvin F. Teachey, Jr., Member problem and shows a very significant load redistribution
taking place due to solar heating.
I commend the authors and NAVSEA for developing a Also, what can the authors tell us regarding their pro-
technical docking tool that all docking activities have gram when the ship in question floats with either a distinct
needed. hog or sag? When docking the Oliver Hazard Perry, we
Norfolk Naval Shipyard's prime objectives for each dock- detected at the time of grounding a sag of about 8 in. in
hag are first, safety; second, facilitate the overhaul work her straight keel length.
package; and third, minimize cost. Over the past ten years Regardless of the predictions that this DRYDOCK pro-

42 DRYDOCK: An Interactive Computer Program


gram has given us, there still exist such a vast amount of by adjusting the pumping schedule to place the dock in a
unknowns for most vessels that shipyards and their dock- hogged condition. I believe this technique is employed in
masters--of limited education and training--must dock certain West European facilities. A similar result can be
under such a variety of conditions of load and floating achieved in a graving dock by adjusting the blocking
dock bouyancy that a conservative approach with ade- height to the desired hogged curve. The authors' com-
quate safety factors is very necessary. ments would be appreciated relative to this approach, and
The severe knuckle loads imposed by surface warships how to input hogging into the DRYDOCK analysis system.
with pronounced stern undercut can be, and have been,
greatly reduced by extending the keel bearing aft of the Robert A. Hohlfelder, Member
knuckle using high keel track towers. Can the authors'
This paper presents procedures which represent a sig-
DRYDOCK program cope with this situation? If so, we
nificant change from traditional methods used by the naval
believe it would be of great help to develop blocking plans
architect to analyze dry dock block loads. The proposed
which greatly reduce keelblock loads. Then, instead of
method is more dependent, however, on detailed knowl-
requiring a 60 000-ton dock to accommodate a 12 000-ton edge of ship geometry, blocking materials and dry dock
warship, a 25 000-ton dock could be used instead, thereby
giving the vessel owner a much greater selection of com- characteristics.
Our shipyard (General Dynamics/Electric Boat Divi-
petitive bidders. sion) currently uses a finite-element structural analysis pro-
gram to model the ship as a beam supported by a number
of elastic springs of varying stiffness, similar to Fig. 1 in
E. K. Moll, Member the paper. The computer model takes into account the
I appreciate the opportunity to comment on the authors' relative stiffness of the submarine hull and of the support
significant contribution to the more precise calculation of blocks. Keelblocks consist of conventional concrete base
the keelblock reactions for unusual ship configurations blocks and wood caps, whereas the bilge blocks are sup-
such as occurs with the DD-963 Class destroyers, the CG- ported on steel I-beams and are a composite of steel, rub-
47 AEGIS Class cruisers, and machinery-aft, fine-lined ber and wood.
R O / R O ' s with stern ramps. In contrast to the author's DRYDOCK program for cal-
In late 1983, the Bath Iron Works designed and analyzed culating block reactions, our computer program accepts
a blocking arrangement to lift a Kidd Class destroyer, the input for the allowable keel and bilge block compressive
USS Scott, at 9515 metric tons (MT). Due to the large stern load and, based on relative block stiffness, calculates soft
overhang and the nonlinear block deflection, we ran sev- cap height adjustments. Bottom hull profile data are fac-
eral iterations of a STRESS model to predict the block tored into the soft cap height adjustment table. This in-
loading more accurately than the results provided by the formation is incorporated in the ship's blocking plan and
traditional trapezoid method. Coincidentally, we made our is used to set the soft caps and to distribute the ship's
work, including full-scale measurements, available to docking weight onto the support blocks.
NAVSEA, its consultant Giannotti Associates (GA), and the Computations have shown that small deviations in soft
Ingalls Shipbuilding Division (ISD) who were for various cap height result in significant changes in block reactions,
reasons comparing actual with predicted block loadings particularly in way of the high rise blocks. As a result, keel
by different methodologies. In the interests of pooling our and bilge blocks were closely monitored by shipyard in-
knowledge, NAVSEA, GA and ISD attended the USS Scott spectors to ensure that soft caps were installed to plan
drydocking and inspected and witnessed the operation of requirements. No attempt was made to measure individual
the No. 3 keelblock load measurement system. This system block pressures once the ship was docked; however, the
utilized hydraulic load cells (four 150-short ton low-profile overall results appeared to be satisfactory based on a visual
hydraulic jacks) integrated with a heavy steel frame top inspection.
and bottom below 20 in. of solid wood blocking covered It is my opinion that the DRYDOCK program could be
with a 4-ft-wide, ~-in.-thick continuous steel plate topped adapted to a particular facility and would provide sub-
with 4-in.-high by 4-ft-square wood caps on 6-ft centers. stantial savings to the shipyard. Conditions leading to im-
The pre- and post-docking vertical distance between the plementation of the DRYDOCK program would be:
frame's upper and lower plates was measured by both a 1. The program should be tailored to smaller, more port-
dial indicator gage and an inside micrometer as a check able, computers. Calculations must be completed in a few
on the inflection point established by the load / deflection hours rather than in a few days.
plot. 2. Ship characteristics and dock characteristics, includ-
As a point of clarification, we have extracted the follow- ing keel and bilge block data, must be readily available.
ing data from our records to correct the BIW data provided 3. Blocking plans must be accurate and must reflect the
in Table 3 and plotted in Fig. 8 of the paper. The only actual block construction.
keelblock load measured by BIW was the No. 3 location 4. Thermal and seismic effects should be factored into
at 306.7 MT compared with our STRESS prediction of 318.5 the program to minimize the potential for overstressing
MT and the DRYDOCK calculated value of 296 MT. The the ship's hull, support blocks or dry dock facility.
closeness of the three results is very encouraging to those
of us seeking improved methods to predict block loading. Assuming that the docking program becomes a reliable
In summary, we fully endorse the general approach pre- tool for analyzing block reactions, the naval architect
sented by this paper. We recommend that instrumentation would be in a better position to:
be used to verify the calculated values of critical block (a) investigate keel and bilge block spacing (along with
loads. The BIW instrumentation is simple, rugged and ec- soft cap area) to reduce the required number of
onomical. No doubt, other measurement systems may be docking positions for hull survey and preservation,
employed for this purpose. and
Also, it should be noted that the floating dock has the (b) recommend various blocking materials that could
flexibility to reduce excessive loading of the after blocks be used as a substitute for wood (because hardwood

DRYDOCK: An Interactive Computer Program 43


t i m b e r of sound quality is b e c o m i n g increasingly docks are r e t a i n e d in about 20 p e r c e n t of the floating docks
difficult to procure). c u r r e n t l y used in Navy facilities.
I want to thank the authors for p r o v i d i n g us with a v e r y To the o t h e r question of discussers H e p b u r n and Luchs,
interesting and timely paper. It is a p p a r e n t that a great the t e n layers of blocking configurations are restricted for
deal of thought and effort w e n t into d e v e l o p i n g the DRY- a single block. T h e r e m a y have b e e n m o r e than one block
D O C K program. to stack up at each n o d e point. W e feel the limitation of
t e n l a y e r s ' r e p r e s e n t a t i o n is m o r e than enough. T h e rec-
o m m e n d a t i o n s to include seismic loading, revised NSTM
997, and m o r e validation are quite welcome. Mr. Teachy's
Authors' Closure suggestion to c o m p a r e the results with the USS Midway
is considered as additional validation. It is worth mention-
The authors wish to thank the discussers for their in- ing that NAVSEA is validating this program; the p r o c e d u r e
terest and r e m a r k s on this paper. The D R Y D O C K pro- will be c o m p l e t e d at the b e g i n n i n g of 1988.
g r a m was d e v e l o p e d to o v e r c o m e the limitations of Mr. Crandall raised a question about the correlation at
existing d r y d o c k analyses and to identify potential prob- o t h e r parts of the ship. This study is in progress at NAV-
lems at the early design stage. Several assumptions and SEA. W e are anxiously awaiting the results. To his ques-
design guidelines can be refined in the future. T h e mod- tions r e g a r d i n g solar heating, distinct hogging or sagging,
ular construction of the D R Y D O C K p r o g r a m was i n t e n d e d these influences n e e d special t r e a t m e n t and a separate
for this purpose. H o w e v e r , the basis for these improve- study. Additional e n d m o m e n t s and block tracing at the
m e n t s should b e studied in detail before they are incor- d e f o r m e d surface is one of the approaches to attack these
p o r a t e d into the D R Y D O C K program. problems. Regarding Mr. Crandall's c o m m e n t on the
Material p r o p e r t i e s of timbers is a major c o n c e r n raised safety factor, the p u r p o s e of this p a p e r is to have a tool to
by several discussers (Mr. G u r k a n and Mr. Kapsilis, Mr. m a k e a good p r e d i c t i o n of block loads. W e leave the safety
Maurer, L C D R H e p b u r n and L T Luchs, and Mr. Mc- factor to the g o v e r n m e n t a g e n c y which relates to this area.
Gruer). This area of study, which should be investigated In response to his last question, about using high keel track
and a p p l i e d to the D R Y D O C K p r o g r a m , involves the ac- towers at the stern, the D R Y D O C K p r o g r a m can handle
curate d e t e r m i n a t i o n of m a t e r i a l p r o p e r t i e s for various this situation. T h e only thing one needs to have is an
types of wood used in the construction of docking blocks. e n g i n e e r i n g j u d g m e n t to simulate the tower stiffness.
W e are aware that the strength of timbers has b e e n the In response to Mr. McGruer's query about ship weight
subject of some r e c e n t study by NAVSEA. T h e question distribution, the nonuniformity of the load distribution can
now is how to i m p l e m e n t the findings into the D R Y D O C K b e r e p r e s e n t e d b e t t e r by increasing the n u m b e r of stations
program. used in the weight distribution curve. Currently, the max-
In response to Mr. Maurer's question, simplification of i m u m n u m b e r of stations for the weight curve is 50. If
ship b o t t o m structure is r e l a t e d to the local deformation. this is not good enough, the user has the option to r e r u n
Keelsons, longitudinal bulkheads, etc. are included in the the S H I P D T p r o g r a m in o r d e r to a d d / r e m o v e concen-
calculation of sectional modulus. T h e flexibility of the bot- t r a t e d loadings.
tom structure is linear s u p e r i m p o s e d on the longitudinal Bath Iron Works i n s t r u m e n t e d keelblocks d u r i n g d r y
b e n d i n g effect. Stanchions are not i n c o r p o r a t e d in the dock operations in o r d e r to m e a s u r e the load on the blocks.
p r o g r a m , but this local structure can b e analyzed ff t h e r e Two ships w e r e e m p l o y e d in the block load m e a s u r e m e n t s :
is a p r o b l e m in the p r e l i m i n a r y study. To Mr. Maurer's First the USS Scott (DD-995), and second the USS Conolly
other question, we d o n ' t suggest considering a block as (DD-979). Measurements w e r e m a d e b y means of a hy-
c o m p l e t e l y u n l o a d e d w h e n the block reaction exceeds the draulic load cell on No. 3 block and crushing load on No.
crushing load. T h e blocks still carry some load, although 1 block. T h e information in this p a p e r is a d o p t e d from
the a m o u n t is uncertain. The elastic modulus of wood and r e f e r e n c e [ 15 ] for USS Conolly. The t e c h n i q u e m e n t i o n e d
crushing load are presently set to constant in the p r o g r a m by Mr. Moll on docking ballast is equivalent to changing
and this can be c h a n g e d at the user's discretion. the stiffness of after blocks. This can be accomplished by
W e agree with L C D R H e p b u r n and L T Luchs that the varying e i t h e r the block height or the block composition.
equivalent plate m o d e l on docks is too simple for the wood Adjusting the dock ballast is a way to achieve this purpose;
floating dock. As m e n t i o n e d in their discussion, the timbers however, we do not r e c o m m e n d the practice since the
used in this d r y dock have b e e n r e l o a d e d for years. W o o d safety of the dock is also one of our concerns.
p r o p e r t i e s are m o r e u n c e r t a i n on the wood dock than the Again, we would like to express our appreciation to all
block timbers. On the other hand, the wood floating d r y discussers for their interest in our p a p e r and for their
dock is phasing out from U.S. Navy facilities. W o o d floating valuable c o m m e n t s and additions.

44 DRYDOCK: An Interactive Computer Program

You might also like