OCR for page 406
th ough a 10 Hz low-pass filter NSEAN uses a
Hottinger Baldwm Messtechmik model Ul, 50 kg
loadcell, sigmal conditioner, ad 16-bit AD cad wi6h
PC for be resists e tests The loadcell, sigmal
c mditioner, ad carriage PC AD card a statically
calibr ted on a Kffmpf ad Remmers precision test
stand to determine the voltage mass relationship Data
requisition is done 6 ough collection of 300 discrete
samples over 10 seconds at 30 H impl fled analog
voltages are converted to frequency (3000~2500Hz) for
transmission to the AD card to reduce sigmal sensitivity
to noise Data is fiitffed th ough a 10 Hz low-pass
filter IIHR uses a Nisshio sham-gags type 20 kg
loadcell, sigmal conditioner, ad 12-bit AD cad wi6h
PC for be resists e tests The loadcell, sigmal
conditioner, ad can iage PC AD card a statically
calibrated on a IIHR test stand to determine be
voltage-mass relationship Data requisition is done
th ough collection of 2000 discrete samples o of 10
seconds at 200 H Data is filtered th ough a 3 H low-
pass filter
Sinkage and trim. Si kage ~ ad him z a
measured wi6h potentiometer-based measurement
systems ad PC data acquisition The potentiometers
sense di placements of the model at the P ad AP
which are conve ted to ~ ad z with equations (2) ad
(3) Data reduction processes a She same as for She
resists e tests DUMB Employs Imear potentiometers,
sigmal conditioners, ad a 16-bit AD card wi6h PC for
the si kage ad trim tests The potentiometers, sigmal
conditioners, ad carriage PC AD card a statically
calibrated on a DTMB test tad to detffmme be
voltage-displaement relationship Data requisition is
done th ough collection of 500 discrete samples over 5
seconds at 100 H Data is filtered Through a 10 H
low-pass filter NSEAN employs rotative
potentiometers, sigmal conditioners, ad a 16bit AD
card with PC for be smkage ad trim tests
Displacement measurements ape made by conversion of
vertical to jugular di placements th ough w ight-
bala ed, mechanical parallelog tams The
potentiometers, sigmal conditioners, ad can iage PC
AD card are tatically c.~ln red on a NSEAN test
stand to detemmine be voltage-displcement
relationship Data acquisition is done 6 ough
collection of 300 discrete samples o of 10 seconds at
30 H Data is filtered th ough a 10 Hz low-pass filter
I HR employs linear potentiometers ad a 12-bit AD
card vifh PC for the smkage ad trim tests The
pOtff tiometers ad can i we PC AD card are statically
calibrated on a IIHR test stand to determine be
voltage-displaement relationship Data requisition is
done th ough collection of 2000 discrete samples over
10 second at 200 H Data is not ffitffed
Wave profile. Wave profiles ~ are measured
with different meiwcement systems ad normah:cd
with model iengh L DTMB utili:oss a waterproof
pencil ad hull-based g id sy tem for the wave profile
tests Wave profiles a marked on the model as it is
tow d 6 ouch be basin Vernier calipers a used to
quatffy be wave profile heights referenced to be full-
load water line NSEAN utilizes a photon Ethic ad
digiti ing measurement system with a hull-based g id
system for the wave profile te Is Data requisition is
done by photographing be wave profile m sections
(20% L) ad digiti ing She negatives with a high-
resolution scanter Wave heights are quantified t x-
stations on be model vifh CAD sof ware IIHR
utilizes adhesive markers, tlex~l:le ruler, level tale,
height gage, a d hull-based g id system for the wave
profile te Is Data requisition is done by tixmg the
adhesive markers t the top of be wave profile at each
x-station The model is removed from the teak ad a
flexible ruler is used to measure be wave profile
dista e along the gi th of the model from be calm
waterline The above two reps a repeated th ee
times The model is Averted ad mo mted on a level
tale ad be average wave height values a remarked
along be gi th of the model from the calm w terline
The height gage is used to measure be wave height z
Far field wave elevations. Fa-field wave
elev lions (~ a measured with eifner caaita - or
servo/aoustic-based measurement systems or PC data
requisition A longitudinalcut medhod is used to
aqui e the far-field data Dnta-redoction is completed
by conversion of longitudmalcut time histories to a
ship coordin te ystem ad Hen norm all ing elevations
with model iengh DTMB .. ~ capacity e-wire
probes sopended fr m a automated 2D havffsmg
system, a 16-bit AD card, ad shore-i aped PC for the
far-field wave elevations tests The probes ad
traversing system are cantilevered from be tank
sid wall on a boom The comity e wi es, 2D
traversing system, ad shore-i aped PC AD card are
st tically calibrated to determine Heir voltage-elevation
relationships Data acquisition is done th ouch
collection of 2000 ad 3000 discrete samples over 20
ad 30 seconds at 100 Hz for Fr=0 28 ad 0 41,
respectively Data is filtered at 10 H Data was
collected t two longitudinal positions: y 0 097 ad
0 324 INSEAN .. . a array of four can ~ it ~ e-wire
probes, moveable slide, 12-bit AD card, ad shore-
based PC for the far-field wave elevation te Is The
probes ad haversing ystem are cantilevered f om be
teak sidewall on a boom The capacity e wires,
moveable slide, ad shore-based PC AD card are
st tically calibrated to determine Heir voltage-elevation
relationships Data acquisition is done th ouch
s
OCR for page 407
collection of 552 discrete samples over 12 seconds at
92 H Data is filtered at 20 Hz Data is collected at a
total of 136 longit dinal cuts Hat are spa d at
fiy=0 0026 betw en She maxim m beam of the model
ad y 0 433 IIHP uses a Kenek servo-type wave
probe ad sigmal conditioner, a Keyence aoustic-type
wave probe ad sigmal conditioner, 2D trazersmg
system, 12-bit AD card, ad shore-i aped PC for She far-
field wave elevation te Is The probes ad traversing
>! rem are cantilevered from the lank sidewall on a I O
m boom The servo ad acoustic probes, sigmal
conditioners, 2D haversing system, ad shore-based PC
AD card are statically calibrated to determine Heir
voltage-elevation relationships Data Requisition is
done th ough collection of 2700 discrete samples over
13 3 seconds at 202 5 H Data is not filtered Data is
collected at a total of 32 longitudmal cuts Hat a
spaced at fly O 01 between the maxim m beam of She
mod I ad the sidewall waw d mpeners (y 0 392)
Near field wave elevations. Near-field wave
elevations (NP are measured with servo-based
measurement >! tems ad PC data Requisition at
DT~LR ad IIHP A trasverse-cut method is used to
aqui e She data at the bow, stern, ad w ke regions
which are inacessible with be longit dinal-cut
medhod Data reduction processes a similar as for the
can iage speed tests DT~LR Employs four DT~LR
whisker probes, sigmal conditioners, 2D traversing
system, 16-bit AD card ad carriage PC for the nea-
field wave elevation tests The whisker probes, sigmal
conditioners, ad carriage PC AD card a statically
calibr ted on be traverse system to determine its
voltage-elevation relationships Data Requisition is
done at 100 H ad the data is filtered t 10 Hz Data is
collected at a tot i of 20 hasverse cuts chat a pa d
at fix=0 0088 in bow ad stern regions Data is
collected with a continuous travffsmg medhod in be y
(transverse) coordinate with fiy~O 0009 IIHP employs
a Kenek servo wave probe, fig al conditioner, 2D
traversing system, 12-bit AD card ad carriage PC for
be near-field wave elevation tests The servo probe,
sig al conditioner, ad carriage PC AD card a
statically calibrated on be traverse >! rem to delerm me
its voltage-elevation relationship Data Requisition is
done th ough collection of 4096 discrete s mples over 9
seconds at 455 H Data is not filtered Data is
collected at a total of 46 transverse cuts chat a spaced
at fix=0 05 in bow ad stern/wake regions Data is
aqui ed with a point-to-point medhod with fly 0 005
between measured pomts
Nominal wake. Nominal wake data (U. V, W.
C; x=0 935) a measured with multihole
probeMifferential pressure trasducer-based
measurement ystffms ad PC data Requisition DT~LR
utilizes a five-hole (3 2 mm tip), bo mdary layer pitot
probe, pitot- tatic probe, five differential pressure
transducers ad sigmal conditioners, 16-bit AD card,
ad cariage PC for be nominal wake tests The five-
hole probe is cabin red before acquisition of the
nommal wake data Pitot tube calibr lion pressure
coefficient ma ices a determined from the calibration
measurements, performed on a calibration rig towed in
calm water with no ship model present The calibration
is expressed m coefficient form as: pitch ankle (Cal
pitch) verms yaw ankles (C yaw), ad axial velocity
(C vel) versus yaw angle Data acquisition for these
experiments was done m a rectagmlar coordinate frame
th ough collection of 500 samples over 5 seconds at
100 H D la is collected at 358 points on 18
horizontal cuts with variable spacing in y ad z Data
reduction is done with be calm water calibr lion
maices NS AN utilizes a pot-side, five-hole,
bo mdary layer pitot probe (3 2 mm tip), pitot-st tic
probe, five differential pressure hasducers ad sigmal
conditioners, 16bit AD card, ad cad i we PC for the
nommal wake tests Th ee calibrations are used for the
nommal wake tests: (1) five-hole pitot probe is
calibrated m be IIHR I 07 m open th o t wind tunnel
(2) differential pressure transducers ad fig 31
conditioners are statically cabin red with water head to
establish the voltage-pressure relationships; ad (3)
calibration is made for be five-hole pitot probe preset
ant le. (of, O. by taking mitial data at each fist lion
with the probe located at s fficient v, z) that mfform-
flow conditions prevail Data acquisition is done
th ough collection of 2000 samples over 2 seconds at
1000 H Data is collected at a total of 32 hori metal
cuts chat a spaced ~0 0025 Transverse spacing of
data is fly 0 0025 Data reduction is done in five steps:
(1) AD card output is statistically 3nahzed; (2) the
average value is converted to mm HO using the
voltage-pressure calibration with Imear mte polation;
(3) velocity vector angles (or, O ad probe calibration
coefficients M, P) are obtained with local Imear
interpolation; (4) correction for five-hole pitot probe
preset ant le from calibration for (by, 0; ad (5) U. V,
W. ad C, are calculated D nsih is calculated as
described for the resists e test HHR utili:D:s She s me
equipment ad procedures as INSEAN except: (1)
HHR uses a propo tionately mailer five-hole pitot
probe which has the same sin: probe tip; (2) HHR
measures starboard-side nominal wake data; ad (3)
HHR collects 1500 samples over 12 seconds at 125 Hz
with a 12bit AD cad
UNCERTAINTY ASSESSMENT
All f ee Institutes follow d ITTC Qualih
Mamal Procedures 4 9-03-01-01 Uncertainh A alysis
6
OCR for page 408
in E D, Uncerteinty Analysis \le~hodoloe -: 4 9-03-01 -
02 k nasnainty Analysis in E D, Guideline for Towing
Tak Tests; ad 4 9-03-01-01 Uncfftamty Analysis,
E ample for R sista e Test The E ample for
R si tan e Test is boded on results from most members
of the 22~3 ITTC Resista e Committee, mcludmg She
present re mlts from NSEAN a d IIBR Bied limits e e
estimated with consideration to sigmfficat elemental
error sources for mdividu d ve fit les, whereas precision
lim its e e e timeted end-to-end for e~pff~mem al re mlts
boded on multiple te Is et the same conditions Total
uncetamties de etimeted with a root-vtm-sqma
RSSI ad normalization with She rage of the result
Tale 2 s mme ices She mcert inty assessment resters
for She overlapmg te Is
Model geometry. Precision limits e e not
considered in She analyses, thus, all of Us is f om She
bied limit For DTMB ad NSEAN, 100% of the bied
limit is associated with the errors in She model
geomet y offsets (x, y, z) which e e e raged over 25 x-
stetions with templates For I HR inn curaies m
loading She model to the correct d aft e co at for 78%
of Us, ad She remeinmg 22% e e associated with She
errors m the model geometry offsets which a
e raged over 31 x-stetions with templates The ettor
in be geometry offsets is sm d l I z O 8 mm), but Us
is relatively high because the model is compe etively
small, he ing 3 5-times less wetted surfa e ea fha
5415 ad 234 A
Caniage speed. Uncertainty in U. et DTMB
is very mall (0 03%) for low ad medi m Fr Bied
errors stffmmmg f om measurement of D ad n a
major contnlxdors to be bied limit Precision limits
contribute mod r uelv ad inmeede with mmeedmg Fr
Uncfftamty in U. et NSEAN is one order of m q,mitud
lager fha et DTMB ad demeedes with mmeedmg Fr
As per DTMB, measurement mcertamty in D ad n e e
sig itt at ad e co at for most of be bied limits
Precision limits a relatively low ad not Fr-dependent
which sugge Is the the d ive motor for NSEAN
towing rank no 2 is stale et all speeds m the test
ma ix Uncertainty in U et I HR is somewhat higher
than et NSEAN due to the I wer rage of speeds for
the smaller model U. also decreases with inmeeding
Fr cd per NSEAN For all Fr, bied limits dommete (82-
94%) U ad decrease with mmeedmg Fr
Measurement of n occurs twice m th data cream ad
acco mts for 95-100% of be bied limits Contributions
from measurement of D me 0 2-3 5% ad errors m the
AD timebede a negligible cd pi DTMB ad
NS AN P, inmeedes with inxeedmg Fr Indicate
red s ed Hi it of be d ive motor to maintain constat
U. for inmeeding speeds, i e, for mmeedmg load on be
motor
Resistance. Uncerteinty in clJdSe et DTMB
is of same order cd values presented by 22~3 ITTC
Resistace Committee, 4 9-03-01-01 Uncerteinty
Analysis, E ample for R sista Test Unce tamty in
the load en calibre ion weight stade d produces high
ad moderate contributions of bied limit to total
ace tamty for Fr=0 10 ad 0 28, re pectively
Precision limit contributions e e rele ively high which
is possibly due to res dual rank motions betw en r drS
Uncfftamty in CT Se et NS AN is mod rued
higher 6ha et DTMB ad decreases with inmeeding Fr
Biedes in be measurement of F. contribute 70%-10%
of B:; for incrceding Fr ad a a fluted to He
static dived calibration w ights ad scatter in the
cabbretion de a ther sig at at factors in Bcr :; e e
Us which conrnlnnei 10%-65% for mmeedmg Fr ad
Us, which contributes 5%-15% for mmeeding Fr The
uncertainties in be water temperature measurement a
net hirable for all Fr cd per DTMB ad IIHR Precision
limit contributions e e relatively high cd per DTMB
which also may be possibly due to re idual tack
motions between runs Uncerteinty in CT Se et I HR
is roughly simile cd for NS AN except for low Fr
where IIHR is 46% lower than NSEAN The buk of
Ucrt5 is bied related with U ad Us conhibutmg 62%-
29% to Bcr :; for mmeeding Fr ad 9%-71% to Bcr :; for
increasing Fr, respectively Effects of scaler m
loadcell calibration dea eccomt for 100% of BE..
Precision limit contributions a rele ively low for all Fr
which may be due to less residual rank motions
between ties cd compar d with DTMB ad NSEAN
Sioliage and trim. Uncertainty in ~ ad z et
DTMB is somewhat le ge for F - O 10 but decreases
sigmfficatly with mmeeding Fr cd the rage of the
measurements inmeede, i e, be potentiometers opera e
further from thei limiting resolution for higher Fr
Bled limit contributions for both vaiahles me mainly
effected by the t alter in She potentiometer calibrations
ad decrease with mmeeding Fr Convert Iy, precision
limit contractions mmeede wi6h mmeeding Fr which
may Indicate elevated cc n we vibration for higher Fr
ad or residual rank motions Uncertainty in a ad z et
INSEAN is le ge for Fr=0 10 but dffreedes
sigmfficatly to values compe able with DTMB wi6h
inmeeding Fr Bled limits e e negligible for bodh
ve iahles ad all Fr due to domma e of precision
limits whose elevated values e e ariboted to redid tat
rank motions between cc ridge r drS Uncertainty in a
ad z et r HR is somewhat le ge for F - O 10 but
decreases sigmfficatly wi6h inmeeding Fr cd per DTMB
ad INSEAN Bled limit conh~botions for a e e high to
mod r.de for Fr=0 10 ad F - O 28/0 41, respectively,
with no Fr-depff dence Bled limit conh~botions for z
e e high et F - O 10 ad decrease wi6h mcreedmg Fr cd
7
OCR for page 409
per DTMB Bias limit magnitudes originate from
scarer m the calibration data Precision limit
conhibutions a mostly high for both variables ad all
Fr a d cau sed by residual tank m otions a d heave/pitch
oscillations at She model mo mt as the model seeks its
hydkody comic equibbri m This 'po poismg' of She
model is likely present at all facilities
Wave profile. Uncertainty in ~ is less 6ha
5% for all facilities ad bodh Fr ad decreases wish
inxeasmg Fr For DTMB ad bodh Fr, bias limits
acomt for 64 5% of Us ad a composed of Form
estimates for marking the wave profile at She exact an-
water mte face (85%) ad measuring the di tan e from
dk ftline to measurement with a vernier Caliph (15%)
Precision limits acco mt for 35 5% of Us ad a based
on 6 ee measurements of She wave profiles ad
estimates of the contact-line msteadiness For
NSEAN precision limits a not considered m She
analysis Contributions to the bias limits are associated
with She uncertainties m the hull-g idline Hick ess
(40%), optical distortion for the camera (50%), scanter
resolution (5%), ad interpolation form (5%) For IIHR
a d both Fr, bias a d precision lim it contributions to Us
are roughly 80% ad 20%, respectively Contributions
to She bias limits a associated with once tamties m the
adhesive marker placement on She hull (50%),
placement of She dk ft ad station lines on the model
(23%), reaplication of She wave profile marks on She
hull ftff She test when the model is removed f om the
teak (23%), a d height-gage readmgs from draftlme to
measurement (4%) Precision limit contributions to Us
a relatively low ad computed for N=3 multiple tests
Far field wave elevrdions. Uncertainty in (~
at DTMB is better 6ha 4% for Fr=0 28 ad 0 41 A
longitudmal cut at y 0 082 is chosen for detailed
assessment of measurement once tamty For bodh Fr,
bias limits are mam contributor to UP ad a
composed mainly of scatter in the calibration data
Precision limits a e timated from N=9 multiple tests
ad mcrease with Fr which may be due to elevated
levels of f e-smfae turbulence with mcreasing Fr ad
"snashot"-like feature of the longitudmal-cut method
Unce t inty m (~ at NSEAN is 3% or better for
Fr=0 28 ad 0 41 Longi5 dinal cuts at y 0 082, 0 172,
0 259, ad 0 347 are chosen for detailed analysis of
measurement mcertainty The myority of UP is bias-
limit related The once tamty in U. acco mts for 70%
of B . wish a 10%-15% conh~bution f om be
once tamty in (x, y) probe position in be test region
L Hart inty m be dista e measurement D betw en the
wave probes ad be FP of be model at 5 0 acco mts
for 5% of B . L Hart inty m the time lag betw en
switch engagement ad data requisition is negligible
Precision limits are estimated withN=10 multiple tests,
contribute moderately to USA, ad decrease wish Fr in
opposition to Hose at DTMB Uncertainty in (~ at
IIHR is better 6ha 3 5% for F - O 28 A longit dinal
cut at y 0 082 is chosen for detailed analysis of
measurement once tamty The myority of UP is bias-
limit related The uncertainty m carriage speed
acounts for 79% of B . with 6%-8% conh~bution
from be ur~rtainty m (x, y) probe positionmg m the
te t region Uncertainty m be dista measurement D
betw en be wave probes ad be FP of be model at t 0
acco mts for 6% of B . Unce tamty m the time lag
betw en switch engagement ad data acquisition is
negligible Precision limits a estimated with N=10
multiple tests ad conhibute sigmificatly to UP as per
DTMB ad NSEAN
Near field wave elevations. For DRAB, low-
turbulence LTR ad high-turbulence HTR (x, y) regions
are identified for det fled analysis in the HTR
precision limit contributions to U5NP are elevated
(85 5%) from large fluctuations in the free smfae ad
air entamrnent mto the flow This produces
comparatively high once tamty in the near-field
measurements For the LTR, contributions of bias ad
precision limits a more balm ed ad be uncertainty
is less than 5% For 6 is case, the sc tter m the whisker
probe calibration governs bias limit magmit de For
I HA LTR (x, y 0 05, 0 07) ad HTR (x, y 1 075, 0)
regions are identified m the wavefield for detailed
analysis Bias limits contribute (25-50%) to URNS
which is mamly (~100%) due to scarer in the servo-
probe calibration data Relatively high PAP is due to
free-smfae turbulence at the multiple-te t conditions
Note chit PAP is th e-times g ester in the HTR fiha
th LTR
Nomdrtal woke. For DUMB, LTR (X, Y.
~0 9346, 0 04, -0 065) ad HTR(X,Y,~O 9346, 0 02,
-O 02) regions are identified m the flowfield for detailed
assessment of be measurement once tamty L bofih
HTR ad LTR bias limit conhibotions to total
uncertamties are dommat for U imd C but somewhat
evenly matched with precision limit contributions for V
ad W. Uncertfi ties in the five-hole calibration
acco mts or 23% of the bias limit wifih the remiinmg
portion doe to probe position uncfftamty in the
flO-.li id ad pressure stahilizition during a can Age
r m Precision limits are estimated from N=10 multiple
tests ad are mod r.te for V, but 15-35% of the total
uncertainty for U. W. ad C which may be a result of
high fi ee-surfiae tmboler~/probe vibration For
LNSEAN, LTR (X, Y. ~0 9346, 0 06, -0 0602) ad
HTR (X, Y. ~0 9346, -0 0025, -0 0602) regions are
identified in the no~fi id for detailed asses merit of fihe
measurement uncertainty Bias ad precision limit
contributions to total uncfftamties of all variables a
8
OCR for page 410
evenly di treated m the LTR ad HTR Bu ad BY
a i fluenced mostly by uncertamties m wind turmel
coefhcients, M ad P. respectively For cros flow
velocities, Bv ad Bw are mamly affected by
uncertainties in velocity vector pitch ad yaw ant les (A
O Contributions f om once tainties m U. to all
variables are less 6ha 3 5%, ad cncenamties m
measurement of pressures at the probe tip a
negligible, except for She center hole where roughly
12% ad 2% contributions to B ad BY are computed
Precision limits a estimated f om N=10 multiple tests
ad are moderate ad high for U ad V, respectively,
but 10-30% for W ad C which may be cased as per
DTMB For IIHR LTR (x, y, ~0 9346, 0 1, -0 01375)
ad HTR (x, y, ~0 9346, 0, -0 02125) regions a
ides tided in the flowtield for detailed assessment of She
measurement uncfftamty Bias limit contributions to
total mcertainties of all variables are dominant in She
LTR ad HTR in She LTR B Ed BY are i fluenced
mo tly by uncertainties m wind turmel coefficients, M
adP,respectively B; ad B a mamlyaffectedby
once tamties in velocity vector pitch ad yaw ant le. (A
O in the HTR By, Bv, Bw, ad B D are mamly
i fleeced by once tamties in She measurement of water
head at She five-hole probe tip since She sin: of She tip is
large with respect to the shear-flow g adients Bv ad
Bw are also still affected sigmfficatly by uncertainty m
the pitch ad yaw angles m She HTR Precision limits
are estimated from N=10 multiple tests ad are ve y
small in relation to She bias limits, which may be due to
lack of free-surface turbulence/pitot-probe vibration
CFD VALIDATION/COM PLEMENTARY CFD
The conditions ad data locations ad densities
(Tale 1) w re selected with consideration to use of
data for CFD validation C vise Fr=0 28) ad flak
Fm0 41) peeds were selected for detailed validation
with most extensive tests for fommer condition No
specific requirement was plac d on experimental
once tamties UD, but rather considered a important
quantity to be estimated at each facility with foal
estimates based on collective results, as discussed
below Additionally, previous experiff e ad
complementary CFD was used for detemmming data
densities
t:l:lhq PARISON OF RESULTS
The focus of be discussions is on be
comparison between in tit tes of facilities, model
geometry, ad overlapmg test results for evaluation of
faility/model geometry ad scale effect biases
Comparisons a made of measurement .! tems ad
procedures ad data (A, B. C) Equation (7)], data
differences D, [equations (8), (10), (12)], data
uncertainties Urx,r,cx ad datadifference
uncertainties [equations (9), (11), (13)] Data
differed es for all variables are computed fter
interpolation of two data sets onto standard dependent
variable values ad subhation Average data
differences are expressed as percentages by non all ing
the average difference with be average rage (all
in titutes) of be ~ an 3} t
Facilities. Faility locations are different as
meekest by latit de, climate, ad zone Yearly average
high ow/dew point temperatures are 66 6°/48 8 /44 4°,
68 0°/51 ~ 5' 8°, ad 59 8°/39 7° 39 5°, respectively, at
A, B. C A ad B have similar mild yearly climates,
whereas C experiences harrhff clim te, e pecially
during the fall ad winter months Zones for A ad B
are on octski Is of large cities, whereas C is in the
center of a small city Such dfffffencer in ambient
conditions a partially acco mted for by using values of
g ad density ad viscosity based on local values of
latit de ad water temper Sure, respectively Water
quality is also different at each facility A ad C ..
local tan water ad B .. . local spring w per; however,
no acco mt is made for water quality since as already
mentioned all mstitutes base values of density ad
viscosity on water temperature only ad fresh w per
values from standard tales
The teaks at each facility have different
dimes sions with the largest to smallest bemg A, B. ad
C, respectively, which affects both blockage ad
residual lank water motions between cariage r ms The
blockage values (ratio of beam/dLaft product ad 1. k
cross-sfftional a a) t A, B. ad C a 0 0017, 0 0042,
ad 0 0055, respectively, which a all nearly at or
below the rff ommff ded maxim m value of 0 005 it is
also recommended chat model length should not be
g eater than lank depth or more than half-as-long as
lank widdh Thus only A satisfies all 6 ee requirements
ad B ad C each violate one E Sects of blockage are
partially acco mted for th ouch conection to can i me
speed Tak sine ad side ad end wall wave d mping
affect residual tack water motions (free ad mb
surface) ad required time intervals betw en can iage
r ms A ad B require 20 minutes betw en can iage
r ms ad C 12 minutes all based on visual observations
In spite of waitmg times, low fiequency motions are
still evident especially m She larger tanks (as mentioned
Hove with regad to large precision limits for smkage
ad trim measurements)
Cap iage speed affects all results, ahhocfh not
always included directly m d la reduction equations
Ahhocfh each instit te measures carriage speed with
similar Instrument lion (encoders ad pulse co mters),
they have different resolutions A has the low st
jugular resolution m thei magnetic encoder (520
9
OCR for page 411
pulses/rev) but the lowest uncertainty, while C has the
highest resolution but the highest uncertainty. The
uncertainty differences are due to the procedures for
transferring pulse count into the data-acquisition PC
(frequency for A and two AD conversions for C) and
the speed range for the uncertainty assessment. Since A
and B operate at nearly twice the carriage speed for a
given Fr, and speed range is the normalizing factor for
determination of Uuc, A and B have less uncertainty in
measured Uc. Uuc can be further reduced through
implementation of a closed-loop feedback system that
constantly assesses and updates Uc with respect to a
desired value. Currently, B is adopting this capability,
however, none of the facilities had this capability
during the overlapping tests. Carriage ride affects
results through carriage vibration; however, such
effects were not considered.
Model geometry. Model geometry offsets,
bow details (leading-edge radius), turbulence
stimulation, surface roughness, and installation also
surely affect all results, although, here again, not
always directly through data-reduction equations. For
each institute, Us is estimated to be 0.5% which is near
the level of Us reported by the participating members of
the 22n~ ITTC uncertainty assessment example for
resistance test. The method for deter ination of the
uncertainties in the offsets is crude as templates are
used at a limited number of stations and error
estimation is tedious, involving moderate and low
accuracy at low- and high-points on the hull surface,
respectively. Accounting for twisting or sagging of a
model in the estimation of the offset errors is a very
difficult, if not impossible task with templates. Also
complicating the issue is changes to the model offsets
over time with changes in ambient temperature and
humidity or by long-ter water-immersion. Leading-
edge radius on each model is 3.2 mm for A and B and
1.6 mm for C. Turbulence stimulation is different at
each facility in terms of tripping location, however,
results in Cider suggest that the turbulence
stimulation was effective for all three models. Model
surfaces were finished using usual procedures at each
facility and assumed hydraulically smooth. All models
were installed according to draft line, which is
presumed the best method for CFD validation pur oses.
However, it was noted by 22n~ ITTC uncertainty
assessment example for resistance test that installation
according to ballast weight reduces uncertainty in
surface area by partially accounting for inaccuracy of
design offsets. Tow points are also different for each
model, but have not yet been compared.
Resistance. For resistance tests, measurement
systems, procedures, and data uncertainty estimates
(Table 2) are similar for all three institutes. Figure 2
displays the results. Figure 2a compares the resistance
data and Figures 2b, c, d compare the data difference
and data-difference uncertainties between institutes AB,
AC, and BC, respectively. Trends in Figure 2a for all
three institutes are typical for high-speed combatant,
i.e., limited humps and hollows for low and medium Fr
and shar ly increasing resistance for high Fr. For low
and medium Fr results for A, B. and C are very
consistent with C usually between A and B. whereas for
high Fr C is consistently lower than A and B.
t
0.0051
0.004
o~ 0.003 _
Inns _
0.00G _
0.0 0.1
(a) Fr
0.0005
m
om 0.0000
(b)
-0.0005 _
0.0
n Inns
. _
om 0.0000
(C)
(d)
---------~------------- DTMB model 5415
. ~ INSEAN model2340A ~-
------------~------------- II H R model 5512
0.2 0.3 0.4 0.5
CRAB
AB
.ooo5o 0
0.0005 ~
.,,,i,,,,i,,,,i,,,,i,...
0.1 0.2 Fr 0 3 0.4 0.5
........... .......... -
AC
U C 1 .1 % C R ~ 3 . %
0.1 0.2 Fr 03 04
0.5
~'~
. . .
U ~ Fr Uo us. 0.5
Fig. 2: Resistance results and data differences.
For AB (Figure 2b), data differences are
oscillatory with Fr and intermittently greater/less than
or within data-difference uncertainty, which suggests
10
OCR for page 412
unaccounted for bias and precision limits. Average
data difference is 2.2%, whereas average data-
difference uncertainty is only 1.5%.
0.0251
0.020
0.015
0.010
0.005
0.000
-nnn.s
(a)
0.010
0.005
~< 0.000
-0. 005
(b)
-0.01 0
0 0 0.1
0.010
0.005 ~
0.000 ~.......
-0.005 ~
(C)
n Gin
-----------~------------- DTMB model 5415 VIA
-----------B------------- INSEAN model2340A
----------~------------- II H R model 551 2
...
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
. . _
0.0 0.1 0.2 0.3
Fr
+U . ~ -
AB . - ~ ~ ..................
~ _
U.U4U
0.035 _
0.030 _
0.025 _
0.020 _
0.015 _
0.010 _
0.005 _
0.000 _
0.4 0.5
-0.005 _
-nn1n
. . ,
................ I
~ .
................ .........
U ~ 1 8 .3 %
02 Fr 03
...................................................................................................
........................................ ................................... .....
.
----------------------- AC---------------------------------------------------------------------------- - -
V. V · V
0 O 0.1 0 2 Fr 03 0.4 0.5
0.010
0.005
0.000
-0. 005
(d)
n Gin.
0.4 0.5
BC - _ .......................................................
_
-_
. K ........................................
. . _ . .
_ .. ............... ..............................................................................................................................
~ -U ~ ~ 16.9 to c=2.6 to
-v.v ~ v
0 0 0.1 0 2 Fr 03 0.4 0.5
Fig. 3: Sin age results and data differences.
For AC (Figure 2c) and low and medium Fr,
trends are similar (i.e., data differences are oscillatory
and intermittently greater/less than or within data-
difference uncertainty), whereas for high Fr the trend is
evident that data difference increases nearly linearly
and is much greater than data-difference uncertainty.
Average data difference and data-difference uncertainty
are 3.4% and 1.1%, respectively. Former is
considerably larger than for AB and latter is somewhat
smaller. Clearly the differences for AC compared to
AB for larger Fr are due to scale effects. Trends for BC
(Figure 2d) are very similar as for AC with average data
differences and data-difference uncertainties of 3.9%
and 1.6%, respectively.
A A A A
-------------~------------- DTMB model 5415
~ INSEAN model 2340A
------------~------------- II H R model 5512
(a)
0.010 _
0.005
0.000 .....
-0.005 ~
(b)
n Gin
, _ . _
0~0 0.1
+U
AB _
0.2 0.3 0.4
Fr
.
.................................................... ............................... .........
.................................................... ....... . .. ....................... ...............................................................................................................................
_UT UTAB 14.1 /0 ~AB_2.5%
is. ~ . ~
0 0 0.1 0 2 Fr 03 0.4 0.5
0.010
0.005
a,< 0.000
-0. 005
(C)
-0.01 0
0.0
0.010
0.005 :
Am 0.000
-0. 005
(d)
a\ a\ ~ ~
1
0.5
0.1
U ...............................
AC . --- - -
~ ~ ,_
............................................................................................................................. . . ... ..= ~.=. = = = = = = .........
U
.U =7.8%
TAC
02 Fr 03
. . .
+U
........................ . .................... . . ...................................................................................................
IAC =5 .5%
0.4
. . . ~
................................................... ...........
. . - . .
................................................... , ........ ~ . . .......................................................................................................................................................
,TBC,
0.2 Fr 03 0.4 0.5
0.5
-ecu ~ u ~
0 0 0.1
Fig. 4: Trim results and data differences.
Results are important in showing that for same
size model (5.72 m) a better estimate for uncertainty in
resistance test is 2.2%, which is considerably larger
than individual facility estimates, especially for
medium and high Fr. Results also show that scale
11
OCR for page 413
effects for 3 m model are significant for Fr>0.26, which
is likely due to differences in wave breaking as will be
discussed later with regard to wave elevation
measurements. For resistance, average facility/model
geometry and scale effect (Fr>0.26) biases are
estimated as 0.5% and 5.8%, respectively, as
summarized in Table 3. Note that averages are based
on all three facilities and Fr ranges as given in Table 3,
and that facility and model geometry biases are
combined as they cannot be separated without use of a
standard model.
Table 3: Summary of facility/model geometry and
scale effect biases.
Result
CR
6
1
(0.28
(0.41
(FF
U
V
W
Fac./model geometry (UF/MG)
AB AC
0.7% 0.9%:
0% 0%
0% 0%l
1.0% 0%
2.6% 1.9%
2.9% 1.1%
0% 0%
0.2% 0%
BC
0%?
0%
0%l
0%
0%
0%
0%
2.3%
Scale (USE)
AC BC
4.9%? 6.7%?
0% 0%
1 1.6%? 12.7%?
0% 0%
0% 0%
0% 0%
0% 0%
0% 0%
13.0% 0% 1 1.7% 0% 0%
I: Fr0.26; ?: Fr>0.33
Sinkage and trim. For sinkage and trim tests,
measurement systems and procedures are similar for all n.nns
three institutes. However, data uncertainties are fairly
large and show some differences (Table 24.
Uncertainties for B at low Fr are very large. Figures 3
and 4 display the results. Figures 3a and 4a compare
the sinkage and trim data and Figures 3b, c, d and 4b, c,
d compares the data difference and data-difference
uncertainties between institutes AB, AC, and BC,
respectively. Trends in Figures 3a and 4a for all three
institutes are also typical for high-speed combatant, i.e.,
increasing sinkage and bow down than up trim for
increasing Fr. In general, results between institutes are
consistent, however, for trim and high Fr, C is
consistently lower than A and B.
For AB (Figures 3b and 4b), data differences
are oscillatory and mostly within the data-difference
uncertainties. Average data differences are 4% and
2.5%, whereas average data-difference uncertainties are
14.1% and 18.3%, respectively, for sinkage and trim.
In this case, data differences are fairly small but with
large data-difference uncertainties. For AC (Figures 3c
and 4c), trend for sinkage is similar as AB although
percentage values are lower, whereas trend for trim is
different for high Fr wherein as with resistance test data
difference increases nearly linearly and is much greater
than data-difference uncertainty. Average data
difference and data-difference uncertainty are 5.5% and
12
7.8%, respectively. Former is larger than for AB and
latter is smaller. Clearly for trim test, as with resistance
test, the differences for AC compared to AB for larger
Fr are due to scale effects. Trends for BC (Figures 3d
and 4d) are very similar as for AC with average data
differences and data-difference uncertainties of 2.6%
and 16.9% and 5.5% and 13.5%, respectively, for
sinkage and trim.
n non
n.n1n
n.nns
-----------~---------- DTMB model 5415
----------~---------- INSEAM model 2340A
IIHR model 5512
-0.005
cat ° °° 0.25 0.50 0.75 1.00
0.010
0.005
Tic 0.000
-0.005
U ~ =5.4% (AB=6.4%
AB
-0.010 _, . . . . , . . . . , . . . . , . .
(b) o so 0.25
0.010 ~
0.50 0.75
x
l l 1
1 .00
-0.005
-0.010 .
(C) ° °°
0.010
0.005
U; =4.8%
AC
0.25
(AC= 3 .3%
0.50 0.75 1.00
~ 0.000 ~ .U~
-0.005 _ (sc
(BC=3 .5%
O.x50 0.75
-0.010 I I
~dy ° °°
Fig. 5: Wave profile results and data differences
(Fr=0.284.
l l l 1
1 .00
Results are important in showing that for same
size model (5.72 m) data differences are fairly small
(i.e., similar level as resistance test) and for low Fr
considerably less than data-difference uncertainties.
All institutes (especially B) need to reduce uncertainties
for low Fr. Results also show that scale effects for 3 m
model are significant for trim and Fr>0.33.
Facility/model geometry biases are not evident for both
sinkage and trim. Scale effect biases are not evident for
sinkage and estimated as 12.2% for trim (Table 34.
OCR for page 414
Wave profile. For wave profile tests,
measurement systems and test procedures are very
different between institutes. However, data
uncertainties have similar values (Table 24. Figures 5
and 6 display the results. Figures 5a and 6a compare
the wave profile data for Fr = 0.28 and 0.41 and Figures
5b, c, d and 6b, c, d compare the data difference and
data-difference uncertainties between institutes AB,
AC, and BC, respectively. Trends in Figures 5a and 6a
for all three institutes are also typical for high-speed
combatant for medium and high Fr, i.e., display bow,
shoulder and stern waves and increasing transverse
wavelength with increasing Fr.
n.n3n
n.n2n
nn1n
-------ale----------- DTMB model 5415
. ~ INSEAM model 2340A
~ IIHR model 5512
-0.010
,^ ~~ 0.00 0.25 0.50 0.75 1.00
Bad x
0.010
0.005
Arc 0.000
-0.005
\ +U;
-------Up
_
~ U =3.1%
(AB
-0.010 _, . . . . . . .
(b) o so 0.25
0.010
0.005
Arc 0.000
-n non
uncertainties. Average data differences are 6.4% and
5.7%, whereas average data-difference uncertainties are
5.4% and 3.1%, respectively for Fr=0.28 and 0.41.
Results for AC (Figures 5c and 6c) and BC (Figures 5d
and 6d) are fairly similar; however, in these cases data
differences are mostly within data-difference
uncertainties.
(a)
n nn
-O. 1 0
-0.20
-0.30
(b)
0.00
-O. 1 0
`-0.20
-0.30
...... " " " " " " .
(C) -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30
Fig. 7: Far-field wave elevation results at Fr=0.28: (a)
DTMB 5415; (b) INSEAN 2340A; (c) IIHR
5512.
-O. 1 0
Anon
(AB= 5. 7%
0.50 0.75 1.00
+U 1 ~
________ (AC -0.10
-U bar
_ . _ _ _ _
U ~ 2 .7 No (AC4 .6%
-0.010
(C) 0.OO 0.25 0.50 0.75 1.00
0.010
0.005
m 0.000
-0.005
-0.010
~ ~ ~ ~ NIB
~dy o so 0.25 0.50 0.75 1.00
Fig. 6: Wave profile results and data differences
(Fr=0.414.
Results between institutes are fairly consistent,
although A seems to show consistently smaller values
especially for shoulder and stern waves.
For AB (Figures 5b and 6b), data differences
are oscillatory and mostly less than data-difference
n On
~ 1
n 1n
~ 1
`-0.20 1
(a) -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30
n nn
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
000 010 020 030 040 050 060 070 080 090 1 00 1 10 1 20 1 30
I . 90
.~ 30
~ 00
-30
~ -60
~ -90
: ~ -120
111111 -15 0
. . . . . . . . . .
-n 2n -n 1 n
-0.30
-...............................................................................
(C) -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30
Fig. 8: Far-field wave elevation differences at
Fr=0.28: (a) DTMB-INSEAN; (b) DTMB-
IIHR; (c) INSEAN-IIHR.
Results are important in showing that a better
estimate for uncertainty in wave profile test is about 5-
6%, which is larger than individual facility estimates,
especially for high Fr. Results also show that scale
effects for the 3 m model are insignificant. Average
facility/model geometry and scale effect biases are
estimated as 0.9% and not evident, respectively (Table
34.
13
OCR for page 415
Wave elevations. For wave elevation tests,
measurement systems and test procedures are different
between institutes. However, data uncertainties have
similar values (Table 24. Figures 7-10 display the
results for the far field tests for Fr=0.28. Comparisons
for the near field tests are not included. Figure 7 and 8
compare the wave pattern data and data differences,
respectively. Figures 9 and 10 compare the data and
data differences for two cuts y=0.324 and 0.082,
respectively. Figures 8a, b, c and 9b, c, d are for AB,
AC, and BC, respectively. Figure 10b is for BC.
Trends in wave patterns for all three institutes are
typical for high-speed combatant, i.e., show diverging
and transverse wave systems originating from bow,
shoulder, and stern. Overall patterns for all three
institutes are very similar, although resolution for A
appears less than that for B and C. Figure 9a shows
fairly large differences between A, B. and C at this
distance from the hull, whereas Figure 10a shows small
differences between B and C close to the hull.
o.oo75 t
n anon
0.0025
0.0000
-0.0025
n anon ~
-0.0075
0.00
(a)
0.004
0.002
m
0.000
-0.002
. - . D. TM.B .mo.del 541 5 interpolate.d -
- INSEAN.model2340Ainter orate
--- - - I- llHR modes 55-1 2 Good - - so - ~ - - - - - - ---- -- -- - --
_ ~ . ~ ~E' >
~ o--. -32-4----- ---------
. .,, i,,,, i,,
0.25 0.50
. ~ ~ ............... ....
................................................................................................................................................... ............... ....
,i,,,,i,,,,i,,,, 1
0.75 1.00 1.25 1.50
X
· FAB
(FF F B 4 .7 % ~ U (FF I
y°.°Oo )0 0.25 0.50 0.75 1.00 1.25 1.50 -n n
0.004
0.002
0.000
-0.002
-0.004
(C) 0.00
0.004
0.002
0.000
-0.002
~ BOA
U (FF =4.4%
AC
~ . . . . . . . . . . .
0.25 0.50
+U(FFAC N A
it_ _ _ _ _ _ _ _ _ _ _ _ _
~ ~ 1 °/n ~
(FFAc
. . . . .
1.25
OFF
AC
. . . . . . . . . .
0.75 1.00
X
+U(FFBC
:: :~~ ~
- ~4 2% F. 4.0% U. FFBC
(d~v.vvO.I 30 0.25 0.50 0.75 1.00 1.25 1. 50
Fig. 9: Far-field wave cut data and differences at
y=0.324, Fr=0.28: (a) interpolated wave cuts;
(b) DTMB-INSEAN; (c) DTMB-IIHR; (d)
INSEAN-IIHR.
Data differences (Figure 8) are largest for the
diverging waves at crests and troughs. For AB, AC,
and BC average data differences are 6.5%, 5.5%, and
2.5%, respectively, whereas average data-difference
uncertainties for AB, AC, and BC are about 4%.
Detailed comparisons at y=0.324 show that data
differences are oscillatory and average data differences
of about 4-5% and data-difference uncertainties of
about 4% for AB, AC, and BC. Here again, largest
differences are at crests and troughs. Trends are similar
at y=0.082, except average data difference is only 0.6%
and data-difference uncertainty is 2.8%. In this case
uncertainty estimates include dependency on x.
0.0075
0.0050
0.0025
0.0000
-0.0025
-0.0050
-0.0075
-0.50 -0.25 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
(a) x
. ~ N
':: ~ ~'
y=0- -08-2~ IN FAN mar 1 994nA
_,,,, i,,,, i,,,, i,,,, i
....................... ...
0.002
c)
l 0.000
-0.002
I ~ ;~__~ ~ ,j' ;~
~ U (FFBC 2 8 /° (FFBC ° 6% -U(FF I
-0.004
(b) -0.50 -0.25 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Fig. 10: Far-field wave cut data and differences at
y=0.082, Fr=0.28: (a) data; (b) INSEAN-IIHR.
)1 _ - U
0.0c
_0.0~
N
-0.04
-0.0E
. ~ L
0.00
-0.02
N
-0.04
-0.06
_ W
: _
Fig. 11: Nominal wake results at Fr=0.28.
Results are important in showing that wave
elevation differences are fairly small and close to the
14
OCR for page 416
uncertainty estimates. Also, scale effects are not
evident. However, visual observation of wave patterns
especially for higher Fr shows differences in wave
breaking, i.e., wave breaking is considerably reduced
for the 3 m model in comparison to the 5.7 m models
presumably due to differences in Weber number and
also some differences between wave breaking patterns
for larger models presumably due to water quality
differences between facilities A and B. Average
facility/model geometry and scale effect biases are
estimated as 1.3% and not evident, respectively (Table
34.
0.00E
_0.0
N
-0.04 ~
n nc
n nc
-n nP
o.oc
N
-in n4
-0.08
-n nP
resolution are apparent. For C, scale effects are not
obvious.
For AB, data differences are less than data
difference uncertainties for U. whereas they are larger
for V and W. However, data-difference uncertainties
(and data differences, except U) are fairly large.
Interestingly, results for AC and BC are similar, i.e.,
scale effects are mostly lost in noise, although pattern
for data differences for both AC and BC are nearly
same, which may be an indication of scale effects.
Results are important in showing that the data
differences and data uncertainty are reasonable close
albeit with fairly large values. All institutes need to
reduce uncertainties. Average facility/model geometry
and scale effect biases are estimated for (U. V, W) as
(0%, 0.8%, 8.2%) and 0%, respectively (Table 34.
Fig. 12: Nominal wake data differences at Fr=0.28.
Nominal wake. For the nominal wake tests,
measurement systems and test procedures are similar
between facilities. However, data uncertainties are
fairly large and show some differences (Table 24.
Figures 11 and 12 display the results for the mean
velocity components (U. V, W) data and data
differences, respectively, for Fr=0.28. Trends for
nominal wake for all three institutes are typical for
high-speed combatant, i.e., relatively thin boundary
layer near keel and free surface and thick boundary
layer near mid girth due to effects of sonar dome
vortex. V and W show influences of sonar some vortex
and upwardfinward stern flow. Flow patterns for all
three institutes are similar, although differences in
-0.06 -
0 0.02 0.04 0.06
y
Fig. 13: DTMB 5415 axial velocity (U) contours and
crossbow vectors for the "/propeller
condition: Fr=0.28, x=0.9603, 436 rpm.
Y 0.4 0.46 0.52 0.58 0.64 0.7 0.76 0.82 0.88 0.94 1
\
Fig. 14: INSEAN 2340A flow mapping at Fr=0.28.
HIGHLIGHTS OF OVERALL TEST PROGRAM
Although the emphasis herein has been on the
overlapping tests, the collaborative effort also consists
of focus studies at each institute. These studies were
15
OCR for page 417
designed to address a variety of important physics and
expand the surface-ship database with quality datasets
and uncertainty assessment. At A, model 5415 was
used in a series of propeller-hull interaction tests.
Shafts and struts were added to the model and data was
obtained with and without the propellers operating.
The measurements include free surface topographies of
the transom wave field, longitudinal wave cuts
(Ratcliffe, 2000), and velocity measurements in the
nominal wake plane and both upstream and
downstream of the operating propulsors. The velocity
measurements were obtained with three-component
laser-doppler velocimetry (LDV). Fig. 13 is a sample
of results for the latter. At B. model 2340A was used
for a comprehensive flow mapping of the boundary
layer and wake flow at eleven cross planes with multi-
hole probes and pressure transducers. The axial
velocity results are plotted in Fig. 14. Results from this
study as well as the overlapping tests and the near field
bow and stern wave elevations from A will be used at
the Gothenburg 2000 workshop on CFD in ship
hydrodynamics. At C, model 5512 is currently being
used for unsteady-flow testing in regular head waves.
Measurements include unsteady forces and moment
with a load cell, far-and near-field wave elevations (Fig.
15) with servo and acoustic probes, and mean and
turbulent flow field with a towed particle-image
velocimetry (PIV) system. The latter measurements are
ongoing and will be archived with the rest of the
unsteady data at I.
Figures 11 and 12 also include comparisons of nominal
wake data for 5512 and pilot and PIV measurement
systems. Data differences are similar to data-difference
uncertainties.
CONCLUSIONS
Results are presented from overlapping towing
tank tests between three institutes for resistance,
sinkage and trim, wave profiles and elevations, and
nominal wake using the same model geometry and
conditions, including rigorous application of standard
uncertainty assessment procedures as per the 1999
ITTC Quality Manual. Two of the institutes used 5.7 m
models whereas the third institute used a smaller 3 m
model. Comparison variables were defined for data-
reduction equations and data differences and data-
difference uncertainties. Detailed descriptions were
provided of facilities, measurement systems, data-
acquisition and reduction procedures, and uncertainty
assessment. Results were discussed with regard to
levels and causes of data differences and data-
difference uncertainties and to estimate facility/model
geometry and scale effect biases.
For same size 5.7 m models, data differences
were in general oscillatory, and in many cases, larger in
magnitude than data-difference uncertainties, which
indicates unaccounted for bias and precision limits and
that current individual facility uncertainty estimates are
often too optimistic. Scale effects for the 3 m model
are only evident for resistance and trim tests for
Fr>0.26 and Fr>0.33, respectively. Facility/model
geometry and scale effect bias are estimated based on
comparisons, as summarized in Table 3. Uncertainty
estimates including such biases may provide better
estimates, especially for use in CFD validation, which
is the recommendation of the present study along with
efforts towards improvement of individual institute
uncertainty estimates. Use of standard models and
current ITTC efforts in providing standard quality
manual procedures for towing tank tests and uncertainty
estimates will also be helpful in this regard.
y-0.2-0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3
0.4
n ~
0.1
O _
)-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3
0.4
n ~
no
0 1
O F.',.i.~ i ~.i~ . ~ ~v~:~.~:.:~.i.~:.:~:.:i.:.i:.>i:.,
(and -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3
O ~,~4, I,.......
(d)-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3
Fig. 15: IIHR 5512 unsteady wave field at four
instances in the encounter period: Fr=0.28,
Ak=0.025, \=4.572 m.
ACKNOWLEDGEMENTS
The research at IIHR and a portion of the
research at INSEAN and at UAH were sponsored by
the Office of Naval Research under Grants N00014-98-
1-0156 and N00014-97-1-0014, respectively, under the
administration of Dr. E.P. Rood. The research at
16
OCR for page 418
DTMB was sponsored by the Office of Naval Research
with 6.2 Funding, administered by Dr. E.P. Rood. The
research at INSEAN was also sponsored by the Italian
Ministry of Transportation and Navigation in the frame
of Research Plan 1997- 1999.
REFERENCES
Avanzini, G., Bennedetti, and Penna, R., 1998,
"Experimental Evaluation of Ship Resistance for
RANS Code Validation," ISOPE '98, Montreal,
Canada, May.
Avanzini, G., Benedetti, and Penna, R., 2000,
"Experimental Evaluation of Ship Resistance for
RANS Code Validation," Journal of Offshore and
Polar Engineering, Vol. 10, No. 1, March 2000
(pp.10-184.
G2K, 2000, MOD
Gui, L., Longo, J., and Stern, F., 1999, "Towing Tank
PIV Measurement System and Data and
Uncertainty Assessment for DTMB Model 5512,"
3r~ International Workshop on PIV, Santa Barbara,
CA, 16-18 September.
Haliday, D. and Resnick, R., 1981, "Fundamentals of
Physics", John Wiley and Sons, New York, 816 pp.
ITTC, 1999, "Report of the Resistance Committee,"
Proceedings International Towing Tank
Conference, Seoul, Korea & Shanghai, China, 5-11
September.
Lee, J., Lee, S.J., and Van, S.H., 1998, "Wind Tunnel
Test on a Double Deck Shaped Ship Model," 3
International Conference on Hydrodynamics,
Seoul, Korea.
Longo, L., Gui., L., Metcalf, B., and Stern, F., 2000,
"Naval Surface Combatant in Regular Head
Waves," abstract submitted 23r~ ONR Symposium
on Naval Hydrodynamics, Vat de Reuil, France,
17-22 September.
Longo, J. and Stern, F., 1998, "Resistance, Sinkage and
Trim, Wave Profile, and Nominal Wake and
Uncertainty Assessment for DTMB Model 5512,"
Proc. 25th ATTC, Iowa City, IA, 24-25 September.
Olivieri, A. and Penna, R., 1999, "Uncertainty
Assessment in Wave Elevation Measurements,"
ISOPE '99, Brest, France, June.
Olivieri, A. and Penna, R., 2000, "Detailed
measurements of wave-pattern and nominal wake
of a fast displacement ship model", AFM
Conference, Montreal, May 2000.
Principles of Naval Architecture, 1967, The Society of
Naval Architects and Marine Engineers, New
York, N.Y., 827 pp.
Ratcliffe, T., 1995, h ~.miV54l S/
Ratcliffe,T., 2000, "An Experimental and
Computational Study of the Effects of Propulsion
on the Free-Surface Flow Astern of Model 5415,"
23r'1 ONR Symposium on Naval Hydrodynamics,
Val de Reuil, France, 17-22 September.
Stern, F., 2000,
Van, S.H., Yim, G.T., Kim, W.J., Kim, D.H., Yoon,
H.S., and Eom, J.Y., 1997, "Measurement of Flows
Around a 3600TEU Container Ship Model,"
Proceedings of the Annual Autumn Meeting,
SNAK, Seoul, pp. 300-304 (in Korean).
Van, S.H., Kim W.J., Kim, D.H., Yim, G.T., Lee, C.J.,
and Eom, J.Y., 1998a, "Flow Measurement Around
a 300K VLCC Model," Proceedings of the Annual
Spring Meeting, SNAK, Ulsan, pp. 185- 188.
Van, S.H., Kim, W.J., Yim, G.T., Kim, D.H., and Lee,
C.J., 1998b, "Experimental Investigation of the
Flow Characteristics Around Practical Hull
Forms," Proceedings 3rd Osaka Colloquium on
Advanced CFD Applications to Ship Flow and Hull
Form Design, Osaka, Japan.
17
OCR for page 419
Table 1: S mmary of ovsrlmpmgteF conditions for DT~, NSEAN, wmdllHR
BXPBR3MBNT DTMB (A) INSBAN (B) 11 B R (C)
MODBL GBOMBTRY Ts~pt lss Ts~pt lss Ts~pt lss
CARRIAGB SPBBD Mcens~ c ~codsv Op~ cM s~codsv Op~ cM s~codsv
RBSISTANCB Lo dceii Lo dceii Lo dceii
Fr 005044 005045 005045
Re OM IMeO7 OM IMeO7 093841eO6
T~ ( Cl 17 8 77 1 75 1
p ~8/m ) 998 6960 997 7700 997 0770
v (mbs) 1 0638e 06 0 9547e 06 0 8M5e 06
p Nim) 0 0738 0 07M 0 0777
DdsdemLy XFr=0013 XF}OOI Fr=001
Model t 1~1 dbc 6 ee 6 ee 6 ee
XFP, AP
SINRAGB TRIM LMwvpols Pot hXpOlS LMwvpols
Fr 005044 005045 005045
Re 0 M 1 MeO7 0 M 1 MeO7 0 86 7 77e O6
T~ (~C) 17 8 77 1 M 5
p ~g/m3) 998 6960 997 7700 997 9400
v (m is) 1 0638e 06 0 9547e 06 0 9708e 06
p Nim) 0 0738 0 07M 0 0733
DdsdemLy XFr=0013 XF~OOI Fr=001
Model t 1~1 dbc 6 ee 6 ee 6 ee
XFP, AP
WAVB PRDF LB Wwp~cli Photo~v yhy 4dh~sbs mavMv
Fr 078,041 078,041 078,041
Re 1 19, 1 75eO7 1 19, 1 75eO7 4 47, 6 54eO6
T~ (~C) 70 6 77 1 18 5
p ~9 m3) 998 1560 997 7700 998 5700
v (m~is) 0 9907e 06 0 9547e 06 1 0448e 06
p Nim) 0 0735 0 07M 0 0733
D dS dem 8y 73pl~ 75Pt 41 pis 40 pis
Model t 1~1 dbc Lxed Lxed Lxed
XFP, AP ( O OM7L, O 00086L), ( O OM7L, O 00086L), (~ 0031L O 00079L),
( O 00054L, O 0083L) ( O 00054L O 0083L) ( O 0015L, O 0079L)
NBAR FI BLD WAVBS Whl Mv probs Ssvvo probs
Fr 078,041 078
Re 1 19, 1 75eO7 3 81eO6
T~ (oC) 70 6 17 5
p ~g/m3) 998 1560 999 5000
v (m is) 0 9907e 06 1 774e 06
p Nim) 0 0735 0 0745
D dS dem 8y 70 OMS, Bx=0 09L, X:=0 0009L 46 sc19, Bx=0 05, X:=0 005
Model t 1~1 dbc Lxed Lxed
XFP, AP ( O OM7L, O 00086L), ( O 0031L, O 00079L)
( O 00054L, O 0083L)
FAR FIBLD WAVBS C yach~ceTvobes C yach~ceTvobes Ssvvo/ccow^cTvobes
Fr 078,041 078,041 078
Re 1 19, 1 75eO7 1 19, 1 75eO7 4 48eO6
T~ ( Cl 70 6 13 3 18 6
p ~8/m ) 998 1560 999 4000 998 5570
v (mbs) 0 9907e 06 1 IM3e 06 1 OMle 06
p Nim) 0 0735 0 0743 0 0737
DdsdemLy 70Ms d:=0097L mdOM4L 136 oc~,Bx=0016,X:=0003 Moc~,Bx=OOOl,X:=OOI
Model t 1~1 dbc Lxed 6t ed Lxed
XFP, AP ( O OM7L, O 00086L), ( O OM7L, O 00086L), ( O 0031L, O 00079L)
( O 00054L, O 0083L) ( O 00054L, O 0083L)
NDMINALWARB J hoieTvobs J hoieTvobs J hoieTvobs
Fr 078,041 078 078,041
Re 1 19, 1 75eO7 1 19eO7 3 83, 5 61e 06
T~ (~C) 70 6 110 17 7
P~8lm3) 998 1560 9996800 9987140
v (m~is) 0 9907e 06 1 76Me 06 1 M73e 06
p Nim) 0 0735 0 0744 0 0744
D dS dem 8y 18 OMS, 358 poM~, variable M OMS, X:=~=0 OM5 77 oc~, X:=~O OM 5
Model t l~lshoc Lxed Lxed Lxed
XFP, AP ( O OM7L, O 00086L), ( O OM7L, O 00086L), ( O 0031L, O 00079L),
( O 00054L, O 0083L) ( O 00054L O 0083L) ( O 0015L, O 0079L)
18
OCR for page 420
Table 2: S mma~y of unce tamties i
or DT=i, 1NSEAN, md IIER ovorl mping tests
DTMB (A) INSBAN (B) IIBB (C)
BBSULT P B P U B P U B P U
S AllFr lOM/o M/o l.VA IOOM M/o O.5V. lOM/o M/o O.5V.
Uc 010 7800/o 22O0/o 0.03~/. 9910/o 090/o 0.14~/. 9420/o 580/o 0.66~/.
U. 028 6820/o 31 80/o 6.63~/. 9370/o 6 30/o 6.13~/. 85 90/o 14 10/o 6.35~/.
Uc 0 41 NA NA NA 99 5M 0 5M 6.1 VA 87 M/o 17 7M 6.1MA
C ~;Cn 0 10 76 30/o 2370/o 1.46~/. 69 40/o 30 60/o 3.64~/. 87 60/o 12 40/o 1.46~/.
C ~;Cn 0 28 45 50/o 54 50/o 6.33~/. 80 O0/o 20 O0/o 6.64~/. 89 20/o 10 80/o 6.63~/.
CTMCR O 41 NA NA NA 66 M/o 33 7M 6.6MA 80 5M 19 5M 6.6MA
7 010 7560/o 2440/o 13.3~/. O0/o 1000/o 43~/. 8220/o 1780/o 4.73~/.
7 028 6840/o 3260/o 5.6~/. O0/o 1000/o 4.71~/. 3040/o 6960/o 1.46~/.
7 041 5630/o 4470/o 3.5~/. O0/o 1000/o 3.63~/. 4280/o 5720/o 6.61~/.
0 10 64 50/o 35 50/o 14.4~/. O0/o 1000/o 33~/. 50 80/o 49 20/o 16.33~/.
028 5470/o 4630/o 3.4~/. O0/o 1000/o 4.76~/. 3610/o 6390/o 1.43~/.
0 41 38 10/o 61 90/o 1.5~/. O0/o 1000/o 6.47~/. 4 10/o 95 90/o 1.76~/.
t 028 6450/o 3550/o 3.53~/. 1000/o O0/o 4.14~/. 8370/o 1630/o 3.43~/.
t 041 6450/o 3550/o 1.44~/. 1000/o O0/o 3.56~/. 8160/o 1840/o 3.66~/.
t w 0 78 14 5M 85 5M 14.MA NA NA NA 75 M/o 74 8M 3.3MA
t w 0 78 56 7M 44 M/o 4.MA NA NA NA 57 M/o 47 M/o 4. 75V.
t w 0 41 M M/o 67 M/o 3.MA NA NA NA NA NA NA
t p 028 76 50/o 23 40/o 3.73~/. 64 90/o 35 10/o 3.46~/. 59 O0/o 41 O0/o 3.43~/.
t p 0 41 66 M/o 33 M/o 3.54V. 78 5M M 5M 3.6MA NA NA NA
U~ 028 7450/o 2650/o 13.5~/. 6040/o 3960/o 6.34~/. 9920/o 080/o 3.11~/.
V P 4350/o 5650/o 6.5~/. 1590/o 8410/o 3.71~/. 9330/o 670/o 3.76~/.
W l 4460/o 3540/o 3.7~/. 8790/o 1210/o 6.46~/. 9920/o 080/o 4.44~/.
Cr T 845M 155M 3.4V. 818M 187M 6.63V. lOM/o M/o 36.3MA
U~ O 6540/o 3440/o 1.6~/. 4780/o 5220/o 6.43~/. 9980/o O20/o 1.36~/.
V T 54M/o 467M 3.MA MM/o 788M 1.47V. 997M OM/o 5.54V.
W 6530/o 3470/o 6.5~/. 7910/o 2090/o 6.65~/. 9990/o 010/o 4.64~/.
Cr 8870/o 2130/o 1.3~/. 7020/o 2980/o 6.11~/. 1000/o O0/o 36.36~/.
U 0 28 42 10/o 57 90/o 3.35~/.
V P 72 40/o 27 60/o 7.73~/.
W l 62 40/o 37 60/o 4.37~/.
V 20 80/o 79 20/o 4.73~/.
v 34 90/o 65 10/o 4.31~/.
w 37 30/o 62 70/o 4.66~/.
nv 13 O0/o 87 O0/o 4.66~/.
~w 30 10/o 69 90/o 5.46~/.
19
OCR for page 421
DISCUSSION
B Beck
University of Michig m, USA
Since you apparently were domg fixed model
tests, was my account taken of the stmding
waves that develop in c towing t mk after c long
day of testing?
AUTHOR'S REPLY
Please see the discussion in section
"COMPARISONS OF RESULTS Facilities "
No special accounting was taken of stmding
waves in the towing tanks due to extended
periods of te tmg Carriage~un intervals were
tw m: mmutes et DTMB md INSE N md
tw 1- e mim tes et IIHR to minimi e free-surface
effects from previous carriage runs Stmding-
wave effects et each facility w re assumed smell
due to effective wave damping
DISCUSSION
L Doctors
The University of New South Wales, AU m clic
I would I ke to con rat late the mthors on then
very carefully conholled experiments, which
must certainly tit te the most precise
experiments on surface ships ever done
It would be mtere ting to estimate She wave
reflection effects due to finite width md finite
depth of She towing t Inks These effects would
be different for all th ee towing t Inks md could
be computed recsorurbly accurately with
Imeari ed waveless tance theo y it might then
be possible to bring the residuary resistmce, for
example, mto better alignment for the thee
models
In Pigme 2(A), w notice c divergence of the
results for the residuary resistance et low froude
mmmber Plotting specific residuary resistance,
rasher f m residuary resistance coefficient,
might Therefore be advised
AUTHOR'S REPLY
Thank you for your kind remarks Differences
md scatter in data for low Pr are largely due to
limited resolution of the mecsur merit System for
low towing speeds md marginally due to
selecti m of cmplffier gain setting which was
optimum for medium md high Pr measurements
The results are presented md discussed es
nondimensional coefhcients to facilitate
comparisons between facilities
DISCUSSION
G Jensen
Her He Ship Model Basin, Germ my
1) How cm the fomm factor be detemmined for c
ship with w tted tr msom according to Prohcskas
method?
2) Did you consider shallow water effects in
crurlyzmg the date or is that part of the "facility
bias"? The INSE N tank is relatively shallow
compared to the model length Therefore, et
high Proude Number some shallow water effect
is expected
AUTHOR'S REPLY
The fomm factor is used only for date reduction es
per ITTC, 1975 The k values are smell, i e,
~0 15 for each facility Shallow water effects
are included in the blockage correction to
carriage speed for each facility th ough the
following equation Tcmurc's fommulc):
~ ~ ~ ~ )
where Am is the midship sectional area of the
model, AT is the sectional area of the t mk, LO i
the length of the model, BT is the t mk widdh, md
Pry is the Pr based on the t mk depthh
DISCUSSION
D Murdoy
Instit te for Marine Development, Canada
Would the mthors please comment on the
possibility that She differences in resistance et
high r En may be associated with the different
trims of the th ee models which, in mm, may be
associated with different tow pomts? Whet w re
the tow points?
Some of the sari tbi it in the date may be due to
differences in the order of test runs Was the
order (for example, low st to highest En,
follow d by highest to low st to fill in gaps)
specified? Whet was the order used for the tests?
OCR for page 422
AUTHOR'S REPLY
The towing pomts w re all close to the model
CG Each facility used (x=0 5, y=0) but towing
height (a) varied somewhat in spite of the
differences in towing height, the sirJcage Ed
trim data was ve y uniform m magnitude for all
models outside of the Fr r mge where scale
effects are import mt, i e, Fr<0 33 11 ~let:3~e, it
is not felt that towing point position was c
signify mt factor in She metsured resi tance
differences, Ether, She differences are I kely due
to wave breaking The order of tests It each
facility m temms of Fr was mdom
DISCUSSION
K I tmur3
Frof Nagasaki institute of Applied Science,
Jcp m
Thanks very much for your presenting She
fundamental st dies on resist mce tests I would
like to ask 3 questions
1) in page 9 of the text, you mentioned blockage
effect, Ed I would like to ask you whether you
conect the effect or not? It is common m Jcp m
to correct it due to the difference of t mk size
2) How do you compensate the effect of cunent
in She tank? In other words, do you measure
water speed direct x?
3) Do you consider She effect of stmdmg wave
of She t Inks used for tests, which may be c msed
by towing c model rep ate Pie m c t Ok?
AUTHOR'S REPLY
The blockage correction is taken into account
th ough Tcmurc's formula (see clove) which
cdju ts She measured towing speed of the model
There is no compensation for the effect of
current in the towing t Inks as only the carriage
speed is measured Ed not also the water speed
directly No special accounting was taken of
stmdmg waves in the towing tucks due to
extended periods of te ting Carriage-run
intervals w re tw nty mmutes at DTAFd Ed
INSEAN Ed tw l-e mmutes at IIHR to
minimize free-surface effects from previous
carriage runs Stmding-wave effects at each
facility were assumed small due to effective
wave damping
Representative terms from entire chapter:
precision limits
Items in cart [0]
