UNH Atlantic Marine Aquaculture Center: Publications

OVERVIEW
The engineering component of the Open Aquaculture Demonstration project has as its primary objective the development of engineering tools for analysis, designing, computer modeling, physical model testing, and full scale evaluation of fish cages and their associated mooring systems. This requires expertise in many aspects of ocean engineering focused on the goal of developing the best tools. A process was engaged that integrates the contributions of the various perspectives to produce the desired results. The remainder of this report demonstrates how the various activities at the University of New Hampshire (UNH) with support from Stommel Fisheries and the Woods Hole Oceanographic Institution (WHOI) have produced this desired result.

There were many goals this past year. The computer modeling was continued with emphasis on improving finite element code for modeling the dynamics of the fish cage in more complex wave and environmental loading. Physical modeling focused on drag force assessment by using scale models in the tow tank. Initial model testing was commenced on the REFA cage, a tension leg cage. The point of investigation was the response of this cage to the wave spectral forcing it would experience at the offshore site. All modeling efforts moved forward into new areas.

A major effort in the past year was the removal, evaluation, and subsequent reinstallation of the north fish cage with new mooring components. Parts of the new mooring components were transducers to sense the line loads and the motion of the cage in response to environmental forcing. This aspect of the work was completed with the support of Stommel Fisheries and investigators from WHOI who were responsible for the complete measurement system, transducers, and telemetry. Their effort and results to date are reported in Irish et al 2000. The remainder of this report documents the efforts of the investigators at UNH.

COMPUTER MODELING

Model Improvements
One of the objectives of the engineering component of the Open Ocean Aquaculture (OOA) project is to develop a methodology of net pen and mooring system computer modeling. To accomplish this objective, the following steps have been performed.

Enhancements and improvements to the finite element program "AquaFE" — a software package developed at UNH to model the dynamic response of partially or completely submerged structures subjected to the complex current and waves loading.
Numerical simulations to analyze the dynamic performance of the currently used Ocean Spar fishcage/mooring system under various environmental loads
Development of the procedure to validate and improve the computer models using physical model tests and in-situ observations of cage/mooring dynamic behavior
Finite element modeling of the tension leg cage "REFA" which is being considered for deployment in the next stage of the project

Enhancements and improvements to the finite element program.
Finite element program "AquaFE" is an advanced computer design and analysis tool to model the dynamic response of partially or completely submerged structures subjected to the complex current and waves loading conditions. Unlike available commercial FEA packages, this program is capable of handling the very large displacements associated with fishcage systems' motion in ocean waves and current. The following enhancements have been introduced in the program:

Development and implementation of the random sea modeling capability,
Introduction of the user-friendly interface to improve the interaction of the core program with pre- and post-processing software,
Improvement of "seabed-mooring gear" interaction modeling by introducing a non-penetration condition into the program,
The problem of adequate modeling of netting in fishcages has been addressed by using the equivalent truss members with geometrical and physical parameters chosen to approximate the dynamic behavior of netting.
Next step in the improvement of the computer modeling of net panels involves development of special net elements — this work is currently underway.

Numerical simulations to analyze the dynamic performance of the currently used Ocean Spar fishcage/mooring system under various environmental loads
Computer simulations have been performed to investigate the dynamic performance of the Ocean Spar offshore fish cage and submerged mooring grid system deployed at an exposed demonstration site in 52 meters of water south of the Isles of Shoals, New Hampshire. The dynamics of the cage motion and mooring line tensions have been analyzed. Both surface and submerged positions of the net pen have been considered. It has been shown that the extreme environmental loading conditions at the demonstration site produce 60% less mooring tension in the case of a submerged cage. According to the analysis, the case when one of four mooring legs becomes disabled will not produce the failure of the mooring system. The results of this study are presented in the paper "Open Ocean Aquaculture Engineering: Numerical Modeling", MTS Journal, Vol.34, No.1

Development of the procedure to validate and improve the computer models using physical model tests and in-situ observations of cage/mooring dynamic behavior
The experimental results obtained for the physical models (scaled as 1/15) in the UNH wave/tow tank have been compared against the results of computer simulations. The first improvement made to the UNH FEA model was to include a bottom conditional in the Fortran code. The code changes were made in the truss element file. A conditional was set such that as a node of an element is equal or less than zero meters, all buoyant, weight and fluid loading forces were set to zero. Since the external forces were eliminated, the inertial and damping forces were also minimized.

Another improvement incorporated into the numerical procedure is the random waves option. The F.E. model creates a random wave signal using a superposition of sinusoidal waves using the random phase method of the form

, (1)

where A_j is the amplitude, k_j is the wave number, s _j is the wave radian frequency (equal to 2p f) and e _j is the random phase. Prior to operating the numerical procedure, the power spectral density of a particular random wave condition is chosen (described as S(f)) for a range of frequencies typically between periods of 2 and 20 seconds. If a superposition of 100 waves, for example, is to be used in the simulation, amplitudes are calculated for a frequency bandwidth (D f) equal to 0.0045 Hz using the expression,

, (2)

where A_j is the amplitude of the individual wave components.

The random phase method is actually deterministic over a period equal to 1/D f when the time series repeats itself. At the end of this period, a new set of random phases needs to be generated to ensure a different surface elevation time series. Currently the F.E. program does not generate its own set of random phases, instead it is performed in preprocessing routines. Since all of the frequency components are included in the period, simulations need not be performed for any time longer. However, the error bounds for the results of a single random wave simulation are so large that little confidence can be associated with them. Additional random wave simulations need to be performed using newly generated phases and the results ensemble averaged (in the frequency domain) to ensure smaller error bounds and therefore higher confidence levels.

Another improvement to the numerical model includes a new element that takes into consideration the shading effects that occur behind up flow net panels. Using this element, certain components of the cage that are known to be in the wake are chosen such that during a numerical simulation, a reduction in velocity can be applied (Figure 1).

Figure 1: Ocean Spar cage model with shaded net, spar and rim elements.

Towing Test Simulations
To compare with the physical model and actual cage tests, towing simulations were performed using the numerical modeling program. Each test was conducted with the ballast weight close-coupled to the spar component of the cage (Figure 2). The tow point was placed 3 meters above the surface interface and the total towline length was 55 meters. The simulation was performed such that 2 meters of freeboard existed. Seven simulations were performed with the current velocity setting ranging between 0.5 and 2 m/s.

Figure 2: FEA model tow test with weight (initial and steady condition).

Free Release Test Simulation
A free release heave experiment simulating the shipyard and physical model tests was also performed using a finite element model representation of the cage using the UNH AquaFE model. The model was constructed using truss elements (Figure 3). The nets were made using the equivalent truss method described in Tsukrov et al., 2000. During the simulations, the water line was placed 2 meters above the static equilibrium position.

Figure 3: AquaFE simulation of the free release test.

Cage and Mooring System Tests in Monochromatic and Random Waves
Additional simulations of the entire fish cage and mooring system are currently being run using the improved AquaFE model. The response of the entire system to both monochromatic and random waves will be investigated. The results will be compared to both the physical modeling techniques and the field tests.

PHYSICAL MODEL TESTING

Overview
As part of the engineering effort of the Open Ocean Aquaculture program at the University of New Hampshire (UNH), physical and modeling methods are currently being developed to design and evaluate fish cage and mooring systems in both waves and currents. Physical models are scaled representations that are built for testing in the 37.5 m x 2.66 m x2.44 m tow/wave basin at the Jere A. Chase Ocean Engineering Laboratory (COEL) at UNH. Results of physical model tests have been used extensively in the past years in the design and analysis of these systems for the Open Ocean Aquaculture Demonstration site south of the Isles of Shoals near Portsmouth NH (Fredriksson et al., 1999 and Fredriksson et al., 2000). Physical model results have also been used in conjunction with the numerical model tests to design systems for the demonstration site.

This section of the progress report describes the component of the engineering tasks associated with the calibration and validation of physical and numerical modeling techniques with field data. It is imperative that these techniques are calibrated and/or validated so that the results are accurate. Inaccurate results can lead to designs that require a high factor of safety which increase component size and weight and therefore deployment and maintenance costs. Three primary sets of field tests are being conducted and the data used to fine-tune the use of the physical and numerical modeling tools. These tests consists of

Drag tests with the Ocean Spar Cage,

Free release tests of the Ocean Spar Cage and

Full cage and mooring system testing in waves and currents.

These field tests are simulated using the physical and numerical models to investigate the accuracy of the methods. The ultimate goal is to understand the dominating physical processes and the limitations of the design tools so that future open ocean aquaculture systems are designed for use that is more efficient.

Towing Tests
During the design and analysis of the mooring system to deploy the Ocean Spar cages at the Isles of Shoals site, the importance of fluid dynamic drag on the cages, particularly the netting, became apparent. Drag forces represent the major physical mechanism by which current and large waves load the cage/mooring system. After deployment, additional questions arose regarding the role of biofouling in increasing fluid dynamic drag. If growth increases the net "blockage" (projected area of net plus growth divided by total area), how much is the drag force increased? In addition, the effects of velocity field decrease acting on structural elements down current from other members (fluid velocity "shadowing") needed to be assessed to effectively model drag force in computer simulations. These issues were addressed in a detailed physical modeling study conducted in the COEL tow tank.

A 1/15.2 scale physical model of the central spar cage manufactured by Ocean Spar Technologies was constructed as described by Palczynski (2000). The model was approximately one meter in diameter, which was large enough for reliable force measurements but was well within the size limitations of the UNH tank. The central spar was made of PVC pipe, while the rim sections were wooden dowels. The basic netting material had a blockage of 5.3%. This was used in single and multiple layers to vary effective blockage. In some experiments, net panels were removed so that the effects of eliminating drag from specific net locations could be quantified. A great deal of care was taken to construct an accurate and robust physical model. Having undergone 216 individual drag test runs, the model is still in excellent shape and will be used in a new study involving wave dynamic response.

The steady drag test program consisted of constant velocity tows in which the drag force on the cage model was measured using a ten-pound capacity Interface Load Cell. Tow speed was measured using two sets of light gates at the beginning and end of the test section. Model tow speeds varied in six evenly spaced increments up to a maximum speed of 0.5 m/s (2 m/s full scale). Tests were done in the normal surface configuration and fully submerged. To assess changes due to an increase in effective blockage from biofouling, one to four layers of netting were used. The effects of net shadowing were studied by conducting additional tows with back panels removed for comparison. Test parameters and drag force results were Froude-scaled to full size values. The low Reynolds number increase in twine drag coefficient was taken into account by adjusting the effective net projected area. This analysis indicated that the three net layer experiments correspond to the full-scale cage when clean of biofouling.

Full scale drag forces acting on the "clean" cage at the surface, Froude-scaled from model results, were 21.4 kN (4800 lbs) at 1 m/s and 77.8 kN (17500 lbs) at 2 m/s. Comparison between surface and submerged test results indicated that wavemaking drag accounts for about 12% of the total drag at a full scale speed of 1 m/s. An increase in blockage of 5.3% to the "clean" cage resulted in an average increase in force of 11% at a full scale speed of 1m/s. Removing sheltered back panels representing 16% of the projected area reduced drag force by only 5% or less. This suggests that down-current net panels are in a shadow region of reduced velocity.

Multiple sets of physical modeling tests were also conducted to simulate the tow tests using the clean net with the ballast weight close-coupled to the spar component of the fish cage. The cage, towing configuration, including the tow point height and towline length was Froude scaled using a 1/15.2 ratio (Figure 4). Seven towing velocities ranging from approximately 0.5 to 2.0 m/s (full scale) were conducted in triplicate. Tow carriage velocity and towing line tension were measured using two sets of light gates and a 45 N load cell, respectively. The results were then Froude scaled to estimate full size values and compared to field test data.

Figure 4: Towing tests being performed with the 1/15.2 scale model.

Free Release Tests
Next, the full-scale free release experiment was simulated using the 1/15.2 scale model. The physical model tests were conducted using multiple net panels to investigate the damping of the system. The free release tests were conducted in the UNH wave/tow basin in front of the observation window on the side of the tank (Figure 5). An optical measurement system was used (Swift et al., 1998, Michelin and Stott, 1996), to measure the motion of the model cage during the free release tests. Four replicates of the heave tests were performed. The results for heave were normalized using the equilibrium position.

Figure 5: Free release test using a 1/15.2 scale physical model.

Cage and Mooring System Tests in Monochromatic and Random Waves
Additional physical model tests were conducted using both monochromatic and random wave conditions. Ten monochromatic wave tests using the periods shown on Table 1 were entered into the wave maker software. An input wave height was entered such that the steepness of the wave (H/L) will be equal to 1/15. Prior calibration tests indicate that for the three longest periods this wave steepness cannot be maintained due to the operating limits of the wave maker. Wave steepness’ for the 1.83, 2.56 and 3.21 second period waves is estimated to be 32.5, 79.3 and 158.5, respectively.

Another set of tests were conducted using a random wave condition estimated to be typical of coastal New Hampshire. The spectrum as shown on Figure 6 was developed using National Data Buoy Center (NDBC) information. This wave condition has a significant wave height (H_s) of 1.21 meters and a peak period (T_p) of 10 seconds. This spectrum will be Froude scaled and entered into the wave maker software using the same 1/15.2 ratio.

Table 1: Monochromatic Wave input parameters.

Period-Full (s) Period-Scaled (s) Length-Scaled (m) Slope H-scaled (cm) H-full (m)

12.515 3.21 13.20 158.5 8.33 1.27

9.981 2.56 9.39 79.3 11.84 1.80

7.135 1.83 5.20 32.3 16.09 2.45

6.238 1.6 4.00 15 26.64 4.05

5.263 1.35 2.84 15 18.96 2.88

4.561 1.17 2.14 15 14.24 2.16

4.016 1.03 1.66 15 11.04 1.68

3.587 0.92 1.32 15 8.81 1.34

3.248 0.833 1.08 15 7.22 1.10

1.949 0.5 0.39 15 2.60 0.40

Figure 6: Estimated wave spectrum for NH (full scale) where H_s = 1.21 meters and T_p= 10 seconds.

The results of the physical model testing in waves will be compared with numerical model results and measurements made in-situ.

FULL SCALE TESTING

Drag Tests
During the redeployment of the fish cage and mooring system in the summer of 2000, a field tow test using the R/V Gulf Challenger was organized using a clean net and the ballast weight attached to the spar. The ballast weight was attached providing a more stable towing situation and close-coupled to prevent fouling in shallow waters. During this test, the total towline length was 55 meters to decrease the influence of the towing vessel wake (Figure 7). A single 20,000 lb load cell was attached just below the A-frame of the R/V Gulf Challenger and connected to the towing lines. It was set to operate continuously recording a point every two seconds.

Figure 7: Towing of the close-coupled clean cage with the R/V Gulf Challenger.

Velocity measurements were made at three locations correlated using a time clock set at Greenwich Mean Time (GMT). The March-McBirney (M-M) current meter was used to collect velocity data in the direction of the tow off the side of the vessel at a depth of 0.3 meters. The M-M current meter was operated at approximately thirty minute intervals collecting five second averages for a duration of one minute. Another set of velocity data was acquired using the shipboard Acoustic Doppler Current Profiler (ADCP). Both water and ship velocity data was recorded from the 5.12 meter depth bin of the ADCP typically within 3 minutes following M-M velocity measurement. A third set of velocity measurements was obtained from an S4 electromagnetic current meter placed inside the towed fish cage prior to the towing operations to investigate net blocking characteristics. The S4 was strung between the top and bottom net panels 3 meters from the surface and 4 meters behind the rim of the Ocean Spar cage (approximately _ the horizontal distance between the rim and the spar). The S4 was set to measure continuously recording 5 second averages. The Captain of the R/V Gulf Challenger, when not maneuvering around lobster pot obstacles changed engine RPM when instructed to obtain velocity and line tension data at different tow velocities.

Free Release Tests
The free release test of the Ocean Spar cage was conducted to investigate the dynamic response of the system. The test was performed in a flooded dry dock at the Portsmouth Naval Shipyard facility in Kittery, ME USA after the cage was refurbished. A crane on a self-powered barge was used during the operations to handle gear and as a work platform (Figure 8). This situation presented a unique opportunity to perform a series of free release tests using the actual Ocean Spar cage in an almost laboratory environment. Once the cage was fully assembled, a Seabird pressure sensor was placed on the bottom net of the cage and set to measure pressure in bursts at a rate of 4 Hertz. A pelican hook and a bridle assembly was connected to the top of the spar and lifted with the crane. When the pelican hook was released, the cage oscillated until it reached its equilibrium position. This test was performed 4 times and the results used to calibrate physical and numerical model methods.

Figure 8: The cage being assembled in the flooded dry dock (prior to release test).

RE-DEPLOYMENT OF THE CAGES & INSTRUMENTATION

Mooring Dynamics Measurements
The objectives of this aspect of the project were to develop a measurement protocol and system and to install this system during the redeployment of the refurbished mooring system. These goals were accomplished in collaboration between UNH and WHOI investigators over the course of the last year. Several meetings that included UNH computer and physical modeling personnel, deployment experts and WHOI researchers were conducted where the pertinent details of the measurement and logistic needs were discussed and defined. Once the needs were defined, the project focused on the hardware specification and fabrication, which was the responsibility WHOI. When the system components were fabricated, their installation was coordinated with the mooring reinstallation. Presently, the load cells are deployed on the northern fish cage and mooring system. Within the next month, accelerometers to measure cage motion and a data telemetry system will be installed.

The first task was to identify critical points in the mooring where dynamic load data was needed for verification of the computer modeling effort. The efforts focused on the need for dynamic data at critical points. Two points in the mooring system were identified. The rope ring where two grid lines, one anchor line, and one lower bridle line are connected was defined as one point. The second location was the upper bridle lines where the cage is attached to the grid buoy. The anticipated loads were defined based on the existing numerical and physical modeling efforts. The maximum loads were anticipated to be on the order of 90 kN (20,000 lbf).

The mooring system was removed in June 2000 according to a plan developed by the UNH / COE faculty and staff in consultation with Stommel Fisheries and the WHOI. The mooring for the north cage was the one that needed attention, and therefore was selected for the measuring system installation. All the mooring components were returned to the NH Port Authority (NHPA) dock in Portsmouth. This was the location for all the subsequent work on the fish cage and mooring system components.

During July, the new and refurbished mooring components were acquired. The fish cage rim sections were tested for leaks. In general, the cage system has no significant structural defects. Only one small porous leak was found in one section end and since the rims are pressurized, no seawater had seeped into the rim section. Evidently, this leak was small, and was an artifact of the fabrication and welding process. Mooring lines were sent to E-Paint in Falmouth to be coated with an anti-biofouling paint. New shackles were acquired. All the gear was checked for wear, and prepared for re-installation. There were some brackets required for the installation of the environmental sensors on the fish cage central spar. The planning and designing of these brackets was done at a meeting at the NHPA with the WHOI. An outside vendor installed the brackets under the guidance of the UNH COE.

In early August the fish cage was reassembled. The dry dock facility at the Portsmouth Navy ShipYard (PNSY) was used to facilitate this task. Three days were required for the cage assembly. The assembled cage was prepared for towing to the offshore site. During the towing a load cell was placed in the towline. Tension data were acquired at various speeds. These data were subsequently used to estimate the drag of the fish cage. The fish cage was temporarily moored offshore for a night waiting for installation in the grid at the site.

A parallel effort to the assembly of the fish cage was the re-installation of the mooring system for the north cage. The F/V NOBSKA operated by Stommel Fisheries from Woods Hole, MA was engaged for this activity. The fishing gear on board the vessel was removed, but the winches and cranes were still in place and used during the mooring work. The mooring installation required two days. Each mooring leg was deployed individually after which the grid was properly tensioned by moving the anchors. This operation was completed as anticipated. The planning by all groups involved enabled the process to occur smoothly. By the end of August, the fish cage was reinstalled in a refurbished grid with the appropriate load cells in place. The final components of the data acquisition and telemetry system that remain will be accomplished in the next month. The next set of tests to be conducted to calibrate/validate the physical and numerical modeling techniques include redeployment of entire fish cage and mooring system with an extensive amount of instrumentation. For modeling purposes, the loads in the mooring system and the motion response of the fish cage will be the focus of the calibration/validation process. Nine load cells have been deployed on the mooring and are recording data. Figure 9 shows the location of three of the load cells and the grid point (three of the load cells are shown on the photograph in the corner of the Figure). Accelerometers are scheduled to for deployment on the Ocean Spar cage to measure the motion response in currents and waves.

Figure 9: Grid corner of the mooring system with three of the nine load cells.

In addition to the deployment of the fish cage and mooring system instrumentation, an oceanographic buoy was installed in November by the monitoring group of the OOA project (Figure 10). WHOI personnel have delivered will a wave rider buoy, which will be deployed by UNH (Figure 10). Data from these buoys will provide the necessary information regarding the environmental forcing at the site.

Figure 10: Oceanographic and wave rider buoys used in the OOA project.

NEW INITIATIVES

Investigation of the REFA tension leg cage
Currently, a new cage system is being investigated for testing at the demonstration site. The system is a different concept from the grid mooring currently deployed. The "Refa" system is a six point, high-tension leg net pen with a small mooring footprint (Figure 11). Computer and physical models of the Refa cage system are being developed to assess force and motion dynamics so that components can be chosen to withstand the high-energy environment of the demonstration site near the Isles of Shoals. One of these critical components is the anchoring system. Another research initiative is the investigation of novel anchoring systems that do not require heavy offshore equipment to deploy but still provide effective holding power.

Figure 11: Refa system currently being investigated.

Finite element modeling of the tension leg cage "Refa" which is considered for adoption in the next stage of the project
The finite element models of the TLC "Refa1800" and 2200 cage systems have been developed (Figure 12). The performance of this cage under typical (Figure 13) and extreme (Figure14) environmental loading has been investigated. Preliminary results show that this cage is able to withstand the loading conditions at the demonstration site provided the appropriate anchor design is chosen. These results will be compared to those of the physical modeling tests and a final assessment made.

Figure 12: AquaFE model of the Refa cage system.

Figure 13: EFA TLC: Deformed shape and maximum tensions under typical loading conditions (wave height=1.2 m, wave length=76 m, current = 0.25 m/s)

Figure 14: EFA TLC: Deformed shape and maximum tensions under extreme loading conditions (wave height=9 m, length=120 m, current changes linearly from 1 m/s on the surface to 0.25 m/s near the bottom)

Refa Cage Physical Modeling

To assess the wave response characteristics of a REFA 1800 and 220 Tension Leg Cages at the Isles of Shoals aquaculture site, a physical model of the cage has been fabricated for testing in the UNH wave tank. Since the entire cage-mooring system dynamics is of interest, the scale ratio of 1/21.3 was based on the ratio of tank depth to site depth. In the planned wave tests, motions will be recorded using a noninvasive optical system, and mooring forces will be measured using a miniature, submersible load cell. A series of single frequency waves will be used to determine the system’s frequency response (response amplitude operator). Random seas will be generated to investigate storm environments and typical daily conditions.

Self Floating Deadweight Anchor Design Concept

The Refa high-tension cage system and the need for a smaller mooring system footprint require large deadweight anchors. These anchors are typically inexpensive to build and require a carefully designed deployment scheme due to their size and weight. Multi-ton anchors require large stable barges with cranes that can lower the anchor to the seabed. These barges and cranes can be expensive and their availability is limited. An alternative method of deploying a deadweight anchor needs to be developed that is cost effective, and can be accomplished with limited local resources. The fundamental concept is to make the anchor floatable so that it can be towed out to the site with a small vessel. To make a deadweight anchor float, it needs to be constructed with a large inner cavity, that when filled with air, makes the anchor buoyant enough to float on the surface. This is the basic premise of a new anchor design currently being investigated.

The design concept being considered is to make a concrete cylinder that is capped on the top and bottom with steel plates. The cavity needs to be large enough to provide the requisite buoyancy, while the total system also needs to provide enough anchoring capacity when deployed. The basic design includes a hole is cut through the top and a pipe is inserted through this hole and secured within a very short distance from the bottom. A smaller hole is cut into the top plate and an air hose is secured to this opening. Figure 15 shows the basic design layout.

Figure 15: Design sketch for the self-floating anchor.

To sink the anchor, water is pumped into the large pipe until the anchor submerges. As the anchor falls, more water flows into the cavity forcing the air out. To prevent the anchor from impacting the seabed, an air bag is attached to the mooring line, some distance from the anchor. The length of the attachment will be shorter than the water depth so that direct impact does not occur. The anchor can then be moved into the final position and be lowered under the control of the air bag. To raise the anchor, air is pumped into the cavity through an air hose forcing the water out through the open pipe. As the water level decreases to a level below the opening of the pipe inside the anchor, the anchor will become buoyant enough to start to rise to the surface. As the anchor rises, expanding air should prevent water from re-entering the cavity.

An experimental model was built to test the concept in a controlled environment. The model was chosen to have a 4 ft -11 in diameter and a 4.5 inch wall thickness, weighing 1585 lbf (dry) so that could be tested safely in the Jere A Chase engineering tank.

Figure 16: Anchor floating in the engineering tank.

In the model design, a centerline wall was added (Figure 19) not only to provide a foundation for a mooring attachment but also to support the top and bottom plates. The centerline wall has a number of air and water ports to ensure easy transfer of air and water between the two halves.

Figure 17: Top view showing the arrangement of the anchor’s concrete walls.

Two water and air holes were cut into the top plate so that there would be one water and air hole for each compartment (Figure 17). The concrete is reinforced with the addition of 80 lb of 3/8 inch rebar and the plates are secured to the ring with _ in threaded rod with hex nuts. The mooring eye is a 1-inch eye nut secured to a threaded rod in the centerline wall. The anchor is designed so that the top plate can be removed and the inner water pipe arrangement can be modified as needed.

Initial tests showed that the anchor was very stable when floating at the surface and could be flooded so that it remained horizontal as it submerged. The more difficult task, recovering the anchor by expelling water with compressed air, required trying various configurations of the internal piping, so that the anchor would rise to the surface and float in a stable configuration.

Figure 18: View of the top plate and its cut openings.

Future work includes additional testing in the engineering tank and a field experiment. The tests will be conducted in the laboratory and once we are satisfied that we can recover the anchor repeatedly from the bottom of the tank, we will proceed to modify the anchor so that we can test the airbag-based deployment procedure in a reasonable depth of water (i.e. 30 to 50 feet of water). After testing the deployment and recovery of the anchor in the ocean environment, the anchor will be modified by welding steel grating on the bottom to increase its "grip" of the bottom. The prototype will then be used to moor a buoy that will impart real world loads on the anchoring system. The performance of the anchor will be observed for at least one month before being recovered. Before being raised back to the surface, the anchor will be dragged to test its holding capacity. This load will simulate the most likely anchor loading scenarios as determined by the engineers working on the fish cage mooring designs.

OPEN OCEAN AQUACULTURE ENGINEERING PUBLICATIONS

Reviewed Journal Publications

Baldwin, K. R. Steen, D. Michelin, E. Muller, P. Lavoie, 2000. Open Aquaculture Engineering: Mooring and Net Pen Deployment. Marine Technology Society Journal. Washington D.C. Vol 34, No. 1 pp 53-58.

Fredriksson, D.W., M.R. Swift, E. Muller, K. Baldwin and B. Celikkol., 2000. Open Ocean Aquaculture Engineering: System Design and Physical Modeling. Marine Technology Society Journal. Washington D.C. Vol 34, No. 1 pp 41-52.

Tsukrov, I., M. Ozbay, D.W. Fredriksson, M.R. Swift, K. Baldwin, B. Celikkol. 2000. Open Aquaculture Engineering: Numerical Modeling. Marine Technology Society Journal. Washington D.C. Vol 34, No. 1 pp 29-40.

Reviewed Conference Publications

Fredriksson, D.W., E. Muller, M.R. Swift, I. Tsukrov and B. Celikkol.. 1999a. Offshore Grid Mooring/Net Pen System: Design and Physical Model Testing, OMAE99-3064. In: Offshore Mechanics and Arctic Engineering ’99. Proceedings of the 18^th International Conference. St. John’s, Newfoundland Canada: ASME.

Fredriksson, D.W., E. Muller, M.R. Swift, I. Tsukrov and B. Celikkol., 1999b. Physical Model Tests of a Gravity-Type Fish Cage with a Single Point, High Tension Mooring, OMAE99-3066, In: Offshore Mechanics and Arctic Engineering ’99. Proceedings of the 18^th International Conference. St. John’s, Newfoundland Canada: ASME.

Tsukrov, I., M Ozbay, D.W. Fredriksson, E. Muller, M.R. Swift, and B. Celikkol. 1999. Offshore Grid Mooring/Net Pen System: Finite Element Analysis, OMAE99-3065. In: Offshore Mechanics and Arctic Engineering ’99. Proceedings of the 18^th International Conference. St. John’s, Newfoundland Canada: ASME.

Department Documents

Fredriksson, D.W., 1998, Design of fish cage mooring system for the aquaculture demonstration site at the Isles of Shoals, System Design Report submitted in partial requirement for the System Design degree program. UNH, Durham, NH.

Muller, E., 1999. Development of an Offshore Aquaculture Site, Master’s Degree Thesis submitted in partial requirement for the Ocean Engineering degree program. UNH, Durham, NH.

Ozbay, M., 1999. Finite element analysis of offshore net pen/mooring systems. Master’s Degree Thesis submitted in partial requirement for the Mechanical Engineering degree program. UNH, Durham, NH., 111p.

Palczynski, M.J., 2000. Fish Cage Physical Modeling. Master’s Degree Thesis submitted in partial requirement for the Ocean Engineering degree program. UNH, Durham, NH., 111p.

Conference Presentations

Several non-archived presentations were given at various aquaculture and engineering related conferences.

Cited References

Charkrabarti, S.K, 1994, "Offshore Structure Modeling". World Scientific Publishing Company, Singapore. 470 p.

Gosz, M., K. Kestler, M.R. Swift and B. Celikkol. 1996. Finite Element Modeling of Submerged Aquaculture Net-pen Systems. In: Open Ocean Aquaculture. Proceedings of an International Conference. May 8 — 10, 1996, Portland, Maine, Marie Polk, Editor. New Hampshire/Maine Sea Grant College Program Rpt. #UNHMP-CP-SG-96-9. pp. 523-554.

Michelin, D and Stott, S., 1996, "Optical Positioning Instrumentation and Evaluation". Ocean Projects Report. University of New Hampshire/University of Maine SeaGrant, University of New Hampshire, Durham, NH. 85 p.

Swift, M. R., Palczynski, M., Kestler, K., Michelin, D., Celikkol, B., and Gosz, M., (1998). "Fish Cage Physical Modeling for Software Development and Design Applications". Nutrition and Technical Devel. of Aquaculture, Proc. of the 26^th U.S.-Japan Aquaculture Symp., UJNR Technical Report No. 26, Hampshire/University of Maine SeaGrant, Kingman Farm, University of New Hampshire, Durham, NH. 199-206.