 wath International Limited has carried out an extensive series
of model tests on two different designs: the 33.5 m 4000 Class and
the 75.5 m long Super Regency car ferry. These model tests along with
theoretical simulations and full scale trials data provide a vital link between
theoretical design calculations and accurate real world performance. Model
tests such as these assure that a Swath International design
will meet it's owners expectations.
 4000 Class Ferry Model Tests
The first SI design to be model tested was the 33.5 m 4000 Class passenger
ferry with a design displacement of 223 tons. A 1/18th scale model was built
in England and used for both the resistance and seakeeping experiments. Two
different strut/hull configurations were evaluated, with the second configuration
receiving the bulk of the testing. These are referred to as the Phase II
tests.
The number 2 tank at the BMT Fluid Mechanics Ltd. facility in Teddington,
England was used for a majority of the calm water experiments. In addition
to bare hull resistance at the design draft of 2.94 m, the following tests
were performed:
-
Trim moment measurement vs. speed
-
Wake survey
-
Flow visualization tests on hull and control surfaces
-
Evaluation of spray rail placement
-
Effect of heavier displacement on resistance
The model was towed from a "trailing link" mechanism. In the Phase I tests
vessel resistance and trim were found to be very sensitive to the towing
method. It is believed that the trailing link method most closely represents
the full scale situation. The towing angle was adjusted for each run to ensure
that the towing force was applied parallel to the shaft line. Resistance
of the model was measured at full scale speeds of 15 to 29 knots. Correlation
between measured resistance and predicted resistance was found to be good.
At each test speed the trim and sinkage were also measured.
Following the calm water tests, seakeeping experiments were carried out in
Tank no. 3 of Westland Aerospace on the Isle of Wight, in England. This 204
m long tank is 4.57 m wide and has a water depth of 1.71 m. Up to 16 channels
of data could be collected. In addition to a full-width model towing carriage
there is a free-to-surge subcarriage which allows for up to 8.0 m of model
fore and aft movement. This enabled the 4000 class model to be towed with
constant thrust when moving through irregular waves. As a result, the
model’s motions were similar to those of a free-running model, and were
considered realistic. Three types of seakeeping tests were performed:
Regular wave testing was carried out at a full-scale speed of 25 knots with
fins fixed and with simulated active control in 1.0 m head waves with periods
ranging from 5 to 13 seconds. One small electric servo motor actuated both
forward and aft sets of control fins in equal and opposite directions. However,
due to some play in the linkages and other sources of lag in the controller,
the model’s fin servo gain was limited to just a small fraction of what
the full scale ship’s control system gain would be. In addition, the
maximum available fin deflection angle on this model was only one-third of
the design deflection angle for the 4000 Class ferry. Nevertheless,
the improved performance provided by the motion control system was quite
apparent. Correlation between measured and predicted motion responses in
regular head waves was considered good.
Irregular or random wave tests were carried out at 25 knots in a simulated
Darbyshire wave spectrum. This is based on Darbyshire’s predictions
for a 100 n. mi. fetch, and is also referred to as the BTTP Standard Inshore
Spectrum. It is used to represent coastal wave conditions and is very narrow
banded with respect to wave frequency. Since the purpose of the irregular
sea runs was to qualitatively observe the behavior of the model in a realistic
seaway, two runs for each test condition were videotaped.
Another series of tests were carried out simulating the 4000 Class
in an evacuation situation, drifting in 3.25 m beam seas at zero speed. The
model was removed from the towing carriage for these tests and was reballasted
to 2 potential evacuation conditions:
-
partially flooded, following damage, and
-
after counterflooding, to facilitate passenger evacuation by reducing heel
and trim.
During these videotaped tests the model was tethered by a several
light control lines which helped maintain the model heading and kept the
model in the center of the tank. Two 1/18th scale 50-man life rafts were
included in the testing to make the simulation more realistic.
Super Regency RO/RO Ferry
 Extensive testing was done on a 1/25th scale model of the Super
Regency design, which has a design speed of 37 knots, displacement of 2300
tons and a draft of 6.05 m. The model was built in the U.K. Tests were performed
at the BMT Fluid Mechanics Ltd. facilities in Teddington, England and at
the Maneuvering and Seakeeping tank of the U. S. Navy’s David Taylor
Model Basin, near Washington, D.C.
All of the tests at BMT were carried out in the 195 m long No. 2 tank, which
has a breadth of 6.0 m and a depth of 2.7 m for 120 m of its length. The
test program included:
-
calm water resistance at design displacement and level static trim
-
wake survey on the port hull at 18 and 37.5 knots
-
self propulsion tests with stock propellers at 5 speeds up to 40 knots
-
measurement of wash wave height at one position off centerline of the ship
-
current force measurements at a full range of headings at a very low speed
-
towed head sea motion measurements in regular waves at 0 and 35 knots with
fixed fins.
 After completing the tests at BMT, the model was shipped to
Martin Automatic in Rockford, Illinois. There it was reoutfitted with a pair
of high-speed electric motors driving the propeller on each hull through
right-angle belt drives. Sealed lead fast-charge cells were added to power
the two motors. Other additions included an actuator for each control surface,
fin control electronics, motion measurement transducers and signal conditioners,
and a clear plastic superstructure. Two separate radio systems were provided:
one to receive operator commands to the model for the throttle, rudder and
fin positions and control modes; and the second to telemeter the measured
data back to shore. The hand held radio control unit allowed all for all
kinds of mixing of channels to provide such features as differential
thrust for low speed maneuvering, and roll-rudder coupling for banking in
a turn.
A schematic of the model’s automatic fin control system is shown in
the accompanying figure. It used a pair of Watson angle and rate sensors
and appropriately scaled gains to reduce both pitch and roll motion. Measured
pitch and roll angles and rates were used in real time to actuate one pair
of fins near the stern at properly scaled rates, and with scaled frequency
response characteristics. The control response rate of a 1/25th scale model
must be 5 times faster than for the actual ship.
When reoutfitting was complete, the model was shipped to Bethesda, MD for
radio-controlled seakeeping tests. The Maneuvering and Seakeeping (MASK)
Facility at the David Taylor Model Basin is 109.7 m long, 73.2 m wide and
6.1 m deep. It is one of the few facilities in the world suitable for testing
a free-running model of the Regency’s size and speed at all headings,
in accurately simulated sea states.
The purpose of the MASK experiments were (1) to evaluate the seakeeping
performance of the Regency ferry in the most realistic manner possible
at the full range of headings in simulated Sea State 5 and 6 conditions
representative of the English Channel, and (2) to evaluate the effectiveness
of an active control system employing a single pair of fins located near
the stern.
The test program consisted primarily of 33 runs with the active control system
functioning at a full-scale speed of 37 knots at 5 headings in a simulated
seaway with a 3.5 m significant wave height and a modal period of 6.0 sec.
Modal period is the period of maximum wave energy. Multiple runs were made
at each heading to improve the statistical reliability of the data. For
comparison purposes, 9 test runs with fixed fins were carried out in the
same sea State 5 conditions at 3 different headings. In addition, 9 test
runs were performed with active control on, at 37 knots in a 5.0 m seaway
having a modal period of 9.0 sec. Most of these runs were at a beam heading.
Two fixed fin runs were also carried out in the same 5.0 m beam seas.
The figure above shows the stages of data acquisition and transmission from
the free-running model to the data analysis computer located alongside the
MASK tank. Eight channels of analog signals from the sensors on the model
were digitized using the transmitter’s A-D card, and then were multiplexed
and transmitted to the shore based receiver via the radio telemetry link.
The onshore unit received the signal and converted the 8 channels of data
back to analog signals. These channels were then filtered using low pass
6-pole Butterworth analog filters. The filtered signals were fed to a PC
equipped with a 12-bit A-D data acquisition board and Labtech Notebook
software. Digitized time histories of each channel were monitored throughout
the test using the graphics capabilities of Labtech Notebook and were
recorded to disk at a sample rate of 30 per second.
Measured motions and accelerations with fixed fins were found to be in good
agreement with predictions for the three headings tested. The measured
significant single-amplitude pitch motions in 3.5 m seas ranged from 0.6
deg. in head seas to 2.7 deg. in stern seas. Sig. roll motions ranged from
0.6 deg. in head seas to 3.9 deg. in beam seas. The highest measured rms
vertical acceleration, 0.080 g’s, occurred at the bow in stern seas.
Highest measured rms lateral acceleration was 0.059 g’s in beam seas.
The effectiveness of the Super Regency’s motion control system,
which is designed to reduce pitch and roll, was evaluated by comparing measured
motions and accelerations with fixed fins to those with active control. There
was found to be essentially no change with active control in the small amount
of pitch motion in sea State 5 head seas, but there was a reduction of about
70% in the measured pitch motion in stern seas. With the active control system
working the amount of pitch motion is nearly the same at all 5 headings tested.
A similarly large decrease in roll motion was measured in beam seas, but
the decrease of rolling in stern seas was also substantial. Active control
also decreased lateral acceleration substantially in beam seas. The measured
effect of control on vertical accelerations was found to vary with longitudinal
location and was heading dependent. Active control had little effect on vertical
accelerations at the model’s longitudinal center of gravity.
These radio-controlled model tests provided motions data which confirm the
predicted excellent seakeeping performance of the Regency class ferry
and the worthwhile improvement provided by the planned active control system.
Further confidence was also gained in SI’s motion prediction computer
program for SWATH ships. |
The Swath
Concept