Operational Evaluation of Stern Boat Deployment Systems

Transcription

1 Operational Evaluation of Stern Boat Deployment Systems R. Sheinberg 1 (FL), C. Cleary 1 (M), F. DeBord 1 (M), W. Thomas 1 (M), R. Bachman 2 (M), J. St. John 3 (M), S. Minnich 2 (M) ABSTRACT Stern-ramp deployment of small boats is an essential feature of all new U.S. Coast Guard cutters. The first new cutter to enter service equipped with a stern-ramp is the USCGC BERTHOLF, the lead Maritime Security Cutter Large (WMSL). During the summer of 2009 a series of sea trials was completed aboard BERTHOLF to define safe operating conditions for stern launch and recovery, develop recommendations for system design improvements, and provide performance data for use during evaluation of future designs. This paper summarizes the results of the boat launch and recovery sea trials and provides recommendations for the design of future systems. Findings are compared to those given in a previous paper presented at the 2003 Annual Meeting (Sheinberg et al 2003) and revised design considerations are presented. These considerations include ramp geometry, limiting ship motions, small boat design features, and special systems such as well water flow management systems and boat capture mechanisms. Finally, the issue of predicting the performance of these systems using physical and numerical modeling techniques is revisited, design criteria are critically assessed, and the importance of operator training is discussed. The views expressed herein are those of the authors and are not to be construed as representing the views or official policy of the Commandant or of the United States Coast Guard. 1. U.S. Coast Guard Surface Forces Logistics Center 2. Naval Surface Warfare Center Carderock Division 3. Science and Technology Corporation Paper No. Sheinberg 1

2 INTRODUCTION Deployment of small boats at sea is a requirement for successful completion of most U.S. Coast Guard missions. Historically, launch and recovery of these small boats have been completed using side davits on large cutters and cranes on patrol boats. Recent designs have incorporated stern ramp launch and recovery systems to replace cranes on patrol boats and augment the side davit systems on cutters. Potential advantages of this approach are: (1) on large cutters, addition of a stern ramp provides a capability for simultaneous operation of more than one small boat; and (2) manpower and time to launch and recover the small boat are believed to be reduced as compared to side davit operations. Differences related to safety and limiting sea states between stern launch and recovery and side davit operations have yet to be determined. To date, stern launch and recovery systems have been incorporated on the Coast Guard WPB 123 ft patrol boats, the WPB 87 ft patrol boats and the Maritime Security Cutter Large (WMSL). In addition, a stern launch and recovery capability is incorporated on the Fast Response Cutter currently under construction. At 418 ft length overall, the WMSL is the only large cutter equipped with a stern ramp. The lead ship in this class, USCGC BERTHOLF (WMSL 750) was delivered in 2008 and a series of boat launch and recovery trials (BLRT) was completed during the summer of The purpose of this paper is to present the findings from the BERTHOLF launch and recovery trials in the context of updating the knowledge base related to system design issues, operational performance, and our understanding of the physics. This information is provided in the form of updates to the 2003 SNAME Annual Meeting Paper by Sheinberg et al Stern Boat Deployment Systems and Operability. Background During the mid-1990s the Coast Guard became interested in stern ramps for launch and recovery of small boats at sea. In order to develop criteria for design and operation of these systems, the U.S. Coast Guard Engineering Logistics Center initiated a project to review the then current state of the art, develop analysis methodologies for evaluating performance, and perform percent time operability (PTO) predictions. That project, as discussed in the reference above included a world-wide survey of existing vessels with stern launch and recovery capabilities, several tank test programs, and sea trials using an offshore oil field service vessel. Those efforts resulted in a set of design criteria, evaluations of the relative operational capabilities for several proposed classes of cutters, and inputs to the design of the WMSL. The performance of stern launch and recovery operations were found to be governed by a complex set of interacting factors related to ship design, small boat design, design of the ramp and associated equipment, and the skill of the small boat coxswains. Key factors include: Stern motions of the ship Ship wake and propulsor wash characteristics Small boat handling characteristics and motions Ramp sill depth, slope, width and shape of the entrance Water behavior within the ramp Availability of an effective boat capture mechanism Skill level of the coxswain In addition, operator judgment in selection of the best ship heading relative to the seas and ship speed were found to be critical, and the interactions between the factors listed above resulted in a very complex decision matrix. For example, head seas are probably optimum with respect to small boat handling, but this heading can result in the largest ship pitch motions and heave at the ramp entrance. Oblique or beam seas typically result in the best ramp entrance motions, but these can be very difficult with respect to handling the small boat. Ship speed variations can be used to minimize ship motions by adjusting the wave encounter period. However, these variations also affect the small boat. Increased speed typically benefits small boat handling, but it can also increase impact accelerations and loads. The project team for the work reported in 2003 also faced a set of issues related to the size of the ship. Stern ramps had historically been installed on patrol-boat-size vessels and the Coast Guard was considering installation on the largest cutters in the fleet. Ship motions, especially ramp entrance motions relative to the small boats were expected to be quite different for the large cutters, and no experience existed for these types of ships. After completing the world-wide survey of existing vessels with stern ramps, two series of model tests were used to evaluate alternate designs and explore the performance trade-offs. During this process the project team developed performance criteria for evaluation of alternate designs and operating strategies. The first set of criteria was related to the safety of the small boat occupants and specified limiting accelerations. These limiting accelerations were based on data collected from public transport studies (Hoberock, 1976) and adopted by the IMO Code of Safety for High-Speed Craft (Werenskoild, 2001). As shown in Table 1, the criteria include three different Safety Levels depending on how personnel occupy the boat. For the previous work, Safety Level 2 was used as the standard. The project team further categorized these criteria as Human and Technical. The Human category was associated with lateral accelerations which were believed to be caused by the operators ability to successfully guide the small boat into the ramp. The Technical category included vertical and longitudinal accelerations which were believed to be caused by ship and small boat design characteristics. Paper No. Sheinberg 2

3 Table 1 Boat Acceleration Limits for Personnel Safety Safety Level Maximum Vertical Acceleration Maximum Horizontal Acceleration 1 Minor Risk of Injury 0.60 g 0.25 g to Person Standing 2 Minor Risk of Injury 0.80 g 0.35 g to Person Sitting and Holding 3 Minor Risk of Serious Injury to Person Sitting and Holding, and No Risk of Injury to Persons Sitting in Safety Harnesses 1.00 g 0.50 g The Coast Guard also had a set of ship motions and acceleration criteria from years of experience with davitlaunched boats based on safety for personnel working on deck at the location of the launch operation. These criteria, which were believed to be equally applicable to stern launch as well as davit (or side) launch, are given in Table 2. Table 2 USCG Boat Launch and Recovery Limiting Cutter Motions at the Launch and Recovery Location Ship Response Maximum Significant Amplitude Roll Angle 8.0 deg Pitch Angle 2.5 deg Vertical Acceleration 0.2 g Lateral Acceleration 0.2 g Finally, a set of criteria were developed specifically for stern launch and recovery to relate ship motions at the stern to small boat accelerations and the expected water behavior at the ramp entrance. These criteria are given in Table 3. The criteria for ramp sill relative vertical motion and sway at the stern were both based on model test results where Safety Level 2 limits given in Table 1 were not exceeded more than 10% of the time. Ramp Availability Time is the average length of time when the water depth at the ramp sill exceeds a specified value, typically the forward draft of the small boat. This criterion is related to the time required for the small boat operator to (1) determine when conditions are suitable for Table 3 USCG Additional Limiting Cutter Motions for Stern Launch and Recovery Ship Response Maximum Significant Amplitude Sill Relative Vertical 7.9 ft (2.4 m) Motion Sway at Stern 4.9 ft (1.5 m) Ramp Availability Time Average 5.0 seconds ramp entry and (2) execute the entry. The value of 5.0 seconds given in Table 3 is a minimum value determined during two sets of sea trials for a smaller ship, and is highly dependent on the specific ramp design, small boat handling characteristics and operator experience. Actual ramp availability times are influenced by ship motions, ramp sill depth, water flow characteristics inside the ramp, and the incoming, radiated and diffracted waves behind the ship. At the conclusion of the project reported in 2003, comparisons were made for alternate ship designs using a parameter called Percent Time Operability or PTO. These PTO predictions were based on motions of the ships calculated using a strip-theory seakeeping program as compared to the criteria given in Tables 2 and 3. These analyses assumed a distribution of wave heights and periods representative of a specific location in the North Atlantic. They did not include consideration for human factors, and ramp availability times were estimated based on relative vertical motions at the stern with no consideration for water flow in the ramp. As such, the PTO predictions given in the 2003 paper are the percentage of total ship operating time where stern launch and recovery could be completed with a 90% probability that small boat accelerations would be less than the Safety Level 2 limits, given the distribution of sea conditions at the specified location. Again, note that human factors were not included, and these were estimated to contribute an additional 10% reduction in the percentage of time when safety limits would not be exceeded. Therefore, if a PTO value of 100% was stated, this would translate to Safety Level 2 accelerations being exceeded 20% of the time. As a result of all of the work described above, several key design and operational recommendations were offered for large ships with stern ramps: 1. Model tests indicated that ramp sill depth is the most important ramp design parameter, and a threshold depth of 0.6m was recommended; 2. An effective water management system in the ramp improved operability by as much as 10%; 3. Ramp entrance shape and width relative to the small boat are important. The ramp should have a funnelshaped entrance, and width should provide a clearance of approximately 0.3m on each side of the boat; 4. The ability of the coxswain is critical for successful recovery. This may be the single most important factor governing operability; and 5. Conflicting information was developed related to the optimum ship heading relative to the seas. Some existing vessels preferred beam seas while the PTO analysis indicated that head seas would be best for a frigate-size ship. This information and other data collected during the project lead to the conclusion that optimum operating conditions are ship-specific and must be determined for each case. DESCRIPTION OF THE SEA TRIALS The 2009 boat launch and recovery sea trials program was completed off of the California Coast in late August. These trials included repeated launch and recovery of the Long Paper No. Sheinberg 3

4 Range Interceptor (LRI) from the USCGC BERTHOLF (WMSL 750). The objectives of these trials were: 1. Quantify the actual safe operating envelope for stern launch and recovery; 2. Assist the crew with determination of optimum operating conditions for launch and recovery in various sea conditions; 3. Identify system modifications that could improve operability; and 4. Develop recommendations for the design and evaluation of stern launch and recovery systems for future large cutters. Testing included variations of ship heading relative to the seas and ship speeds for a range of sea states, to quantify launch and recovery performance and to collect engineering data for future designs. In addition, detailed measurements of ship motions characteristics, stern ramp hydrodynamics and small boat ramp impact loads were completed. Figure 1 is a photograph of BERTHOLF and principal characteristics are given in Table 4. BERTHOLF is the lead ship in the WMSL Class, which are the largest cutters in the Coast Guard fleet. These vessels are designed for extended long-range patrols and missions including maritime law enforcement, search and rescue and homeland security. Completion of these missions requires launch and recovery of small boats, and the stern ramp was included to facilitate these operations. that the ramp can accommodate various size boats. The ramp bottom consists of a grating supported by the ship s transverse frames, and some water management is provided by the open area beneath this grating. Support rails are Ultra Poly mounted rigidly to the flanges of longitudinal steel beams. A boat capture mechanism is included in the ramp as shown at the bottom of the photo on the first page of this paper. This mechanism consists of a capture line across the ramp (just visible above the safety net in the photo) that falls over a horn on the small boat when the boat bow passes beneath the line. The capture line is attached to a carriage that moves the boat up the ramp after it is successfully captured. A safety net is also included forward of the capture line to prevent the boat from running too far up the ramp in the event of a large wave surge. Table 4 USCGC BERTHOLF Principal Characteristics Characteristic Value Units Length Overall 418 (127) ft (m) Beam 54 (16.4) ft (m) Draft 21 (6.4) ft (m) Full Load Displacement 4,500 (4,570) Maximum Sustained Speed 28 (14.4) Table 5 WMSL Stern Ramp Characteristics LT (Tonnes) s.w. knots (m/sec) Characteristic Value Units Length (horizontal) 41 (12.5) ft (m) Width (3.85) ft (m) Entrance Width 18.0 (5.5) ft (m) Slope 14 deg Sill Depth 2.3 (0.64) ft (m) Figure 1 USCGC BERTHOLF (WMSL 750) Ramp geometry characteristics are given in Table 5 and Figure 2 shows the interior of the ramp with the stern doors closed. Key features of the ramp are a slope of 14 degrees which is near the minimum required for reliable selflaunch, and a width of 12.5 feet, which provides slightly more than 1 foot of clearance on each side of the small boat. The depth of the ramp entrance at the transom (sill depth) is 2.3 ft below the nominal stern water line. The ramp entrance at the stern doors is wider than the ramp interior, providing a funnel shape to help guide the boat into the ramp. The two interior parallel boat support rails in the bottom of the ramp also have a funnel shape at the stern. Two additional boat support rails are provided such Figure 2 WMSL Ramp Interior The LRI is the largest Coast Guard Rigid Hull Inflatable Boat (RHIB) currently planned to operate from stern ramps on cutters. It is designed for high speeds and endurance, Paper No. Sheinberg 4

5 and is intended for operations at significant distances from the cutter. Principal characteristics are given in Table 6, and Figure 3 is a photograph of the boat. This figure shows the bow horn used for capture just above the orange collar at the tip of the bow. The boat is powered by diesel engines driving twin water jets. Engine cooling systems permit starting five minutes prior to launch. Table 6 LRI Principal Characteristics Characteristic Value Units Length Overall 35.0 (10.7) ft (m) Maximum Beam 10.5 (3.2) ft (m) Full Load Displacement 24,000 (10,890) lbs (kg) s.w. Deadrise Angle 25 degs Maximum Speed 45 (23) knots (m/sec) Figure 3 Long Range Interceptor (LRI) Trials Instrumentation A major effort was undertaken to insure that instrumentation was provided that would satisfy all project objectives and provide engineering data not available from model tests, numerical predictions or trials with smaller vessels. Two separate instrumentation systems were installed for these trials, one onboard BERTHOLF and a separate system onboard the LRI. In addition to the measurements necessary to quantify operability as discussed above, significant efforts were made to collect design data for ramp water-flow dynamics and small boat impact loads. These two areas are considered to be important for improving system capabilities in the future and very little data was available from the previous work. The shipboard measurements consisted of ship motions, ship operating parameters, the environment, and ramp availability sensors. An Ixsea inertial measurement system (INS) provided 6 Degrees of Freedom measurements in the vicinity of the ship s center of gravity. Tri-axial accelerometers were located on either side of the stern ramp on the transom and in the pilot house. Ship systems, such as rudder angles, propeller pitch, shaft speed, and ship speed were measured directly from installed sensors and/or accessed from ship s interfaces. Environmental, and other external, measurements were recorded for wind speed and direction, ship s position, and speed over ground. A Datawell directional sensing wave buoy was deployed to characterize the seas. Given the complex environment of the changing sill depth, with potential for spray, the ramp availability measurements employed redundant sensors. Water level above the sill was measured by Global Water pressurebased water level sensors installed on the outboard side of the inboard skids, beneath the flange so as to be sheltered from the potential impact of the boat hull. One each was installed port and starboard as far aft as possible at the end of the skids. These sensors were approximately two feet forward of the sill and the lower most part of the ramp. The data were subsequently translated to the sill. Four additional water level sensors were installed at sequential distances from the sill into the stern well, as seen in Figures 4a and 4b. All water level sensors were horizontally mounted on top of the grating facing aft, with the exception of the third one from the sill. This sensor was mounted vertically facing down into the water drainage compartment. Water velocity sensors were installed to measure water velocity flowing into and out of the stern well, as seen in Figure 4b. These measurements had a dual function of quantifying the dynamic head component of the water level measured by the pressure sensors, and combined with the water level measurements, quantifying flow rate. Two types of flow sensors were used. A propeller type, Swoffer, and a paddlewheel type, Airmar, were mounted in pairs. For each pair, one sensor faced forward and the other aft. These four sensors were mounted on the inboard port side skid forward and aft of the knuckle. Figure 4a - Ramp Availability and Water Management Sensor Locations Paper No. Sheinberg 5

6 Water level at the sill was also measured by an overhead acoustic sensor. A Senix Tough-Sonic distance measurement sensor, mounted on a beam, was oriented to look-down at the centerline of the sill. A second Senix was mounted over the well forward of the primary Senix. These two sensors can be seen in Figure 4a. A third Senix was mounted approximately six to eight feet aft of the transom. The overhead sensors were mounted such that they were easily retracted during boat operations. The trials data collection system and video recorder station were established in the boat shop work space. Three fixed cameras were mounted in the ramp area of the cutter. The first camera was placed beneath the flight deck close to centerline. It faced aft and was pointed down into the positions were also measured by two potentiometers mounted on the operator s control levers, one for starboard and one for port. Jet drive position for steering was measured using a single potentiometer attached to the connecting rod in the engine compartment. The entire measurement and data acquisition system was powered by six 12-volt marine batteries distributed throughout the boat such that they did not affect the center of gravity location. A video system was used to capture the behavior of the LRI and the orientation of the boat relative to the cutter during testing. A camera was installed on the front of the cabin of the LRI facing forward to capture the coxswains view, and a second camera was mounted to record action at the helmsman s control station. With both cameras installed, a video mixer combined real time motions of the LRI and the helmsman s actions as a picture-in-picture recorded on a digital video recorder. Impact load measurements on the LRI were accomplished by strain gauging the framing structure at the bow. Specifically, three stringers supporting the shell plating on either side of the keel were fitted with shear strain gauges placed at opposite ends of each stringer segment spanning one frame. This approach was used for all sections between frames 2 and 6, except for the outboard-most starboard stringer at frame 5 where a welded foundation would significantly limit the response. The keel was instrumented in a similar manner. The general arrangement of the strain gauges is illustrated in Figure 5. Figure 4b - Ramp Availability and Water Management Sensor Locations ramp. The second camera was also on the centerline, but positioned low in the well looking aft. The third camera was located on the transom, looking athwartships over the sill. An additional roaming camera was available to document other views as necessary. The types of measurements completed onboard the LRI are summarized in Table 7. The installation required that the data acquisition system be mounted in an equipment rack shock mounted and bolted to the cabin floor. The rack supported a National Instruments data acquisition system with signal conditioning, and power supplies for all sensors. A Global Positioning System (GPS) receiver and the Inertial Measurement System were installed in the cabin near the center of gravity location. The inertial measurement package had its own internal GPS receiver with a smaller antenna but it was also coupled to a Novatel GPS. This two-antenna system increased GPS accuracy. A separate triaxial accelerometer was mounted on the forward-most bulkhead in the forward compartment to measure translational accelerations due to impacts. Waterjet bucket position was measured using two LVDT s. The LVDT s were connected to the starboard and port bucket control arms in the machinery compartment. Throttle Table 7 - Summary of LRI Measurements General Measurements Date and Time Boat Position Boat Speed (knot) Boat Heading (ºT) Motions Measurements Triaxial Accelerations at the boat CG (g SSA) Triaxial Accelerations at the bow (g SSA) Roll (º SSA) Pitch (º SSA) Heave (ft SSA) Impact Loading on Boat Bottom Structure As shown in Figure 6, the gauges were placed approximately one inch above the shell plating, close to the combined neutral axis of the bottom plating and longitudinal stringer. The locations of highest shear strain were identified from detailed finite element analysis of the frames. The gauges were located close to, but just far enough from, the web frames and bulkheads (and limber holes, if present) to obtain a reasonable response on the order of an inch. Impact loads were related to the difference in strain between the gauges on either end of a stringer in a frame bay using results of finite element analyses combined with a physical calibration. Paper No. Sheinberg 6

7 Table 8 Desired Test Matrix Parameter Variations Significant Wave Height (ft) 0, 3, 6, 9 and 12 Ship Speed (knots) 5 and 8 Ship Heading Relative to Waves (deg) 0, 30, 60, 90, 120, 150 and 180 Boat Relative Speed (knots) 4, 6 and 8 Boat Heading Relative to 0, leeward, windward Ship been reached. This process was repeated as each different sea condition was encountered. Figure 5 Plan View of LRI Bow Showing Arrangement of Impact Load Strain Gauges on Longitudinal Bottom Stringers A second modification to the original test plan intended to increase productivity was inclusion of touch-and-go tests. These tests included a recovery without capture followed by an immediate launch after the small boat entered the ramp and reached a position where capture would have been possible. Cycle times for these tests were approximately 5 minutes as compared to 15 to 20 minutes for a recovery with capture and a launch. For most test conditions (sea state, ship speed, ship relative heading) testing included a launch, four touch-and-go tests and a recovery with capture. In addition to launch and recovery tests, objectives required that ship motions and stern well water-flow characteristics be measured at speeds and headings similar to the launch and recovery tests. For these tests, data was collected while the ship held a constant speed and heading relative to the seas for approximately thirty minutes. In most cases, the wave buoy was deployed at the beginning of these tests and recovered when they were complete. Figure 6 - Shear Strain Gage Installation on Bottom Longitudinals for Impact Loads Test Procedures Satisfying project objectives required completion of launch and recovery operations for a range of sea states, ship headings relative to the seas, ship speeds, and small boat approach strategies. The desired ranges of variations for these parameters are given in Table 8. As shown in the table, a complete matrix of these conditions would require completion of 630 launch and recovery evolutions. In addition, since statistical descriptions of operability were the desired outputs, repeat tests at each condition were necessary. Due to limited availability of ship time, an alternate approach was developed for selecting specific test conditions. Based on the ship s previous experience, testing for a given sea state began with conditions believed to be somewhat difficult, but not near operating limits. Subsequent tests were selected with incrementally increasing difficulty until measured results or crew observations indicated that the safe operating limit had Test Conditions During the five-day trials period, 22 stern well water flow observation tests and 164 launch and recovery tests were completed. Launch and recovery tests included 33 individual launches, 37 individual recoveries with capture, and 94 touchand go tests. Sea conditions ranged from significant wave heights of 2.5 ft (Sea State 3) to 8.4 ft (low Sea State 5). A summary listing of all tests is given in Table 9. During the trials period sea conditions included wind waves and a background swell, typically from different directions. This condition was especially prevalent during the first two days of the trials, complicating data analysis and interpretation of results. For all results included below, significant wave height was calculated based on the total energy of the combined seas, and the wave directions were defined as the direction of maximum wave energy. This interpretation of the wave direction data resulted in no data collected at some targeted ship headings even though they were believed to be completed as testing progressed. Specifically, relative headings of 0, 30 and 90 degrees were not achieved in Sea State 3, even though tests were conducted with these apparent wave directions. For testing in Sea State 4, headings of 90, 120, and 150 degrees were intentionally omitted based on the crews experience and results of the Sea State 3 tests. Paper No. Sheinberg 7

8 Table 9 Summary List of Launch and Recovery Tests Sea State Ship Relative Heading Ship Speed Number of Tests (deg) (knots) Total Total Grand Total 164 TRIALS RESULTS Results from the sea trials include data related to the operability of the stern launch and recovery system onboard BERTHOLF as a function of sea state, ship heading and ship speed, as well as data sets related to the factors discussed above that affect operability. In addition, several special data sets were collected related to specific ship and boat design issues. Due to the Coast Guard s law enforcement and homeland security missions, operational capabilities such as definition of conditions suitable for small boat operations cannot be released to the public. Therefore, the presentation of results will be focused on providing insight into the factors affecting operability, refinement of design criteria for future vessels, and the special engineering data collection efforts. At the end of this section, several examples of operability results will be presented without providing actual percentages, to illustrate the trends in performance with sea state, heading and speed. Small Boat Accelerations and Motions With the exception of limiting ship motions for personnel working on deck, all of the previously developed criteria for launch and recovery safety are based on small boat impact accelerations. Even the criteria for ship stern motions were derived from the small boat accelerations measured during model tests. Therefore, small boat motions and accelerations will be discussed first. Table 10 provides a summary of small boat accelerations and motions for all launch and recovery tests completed during the trials. The values listed as Max SSA are maximum significant single amplitudes calculated as two times the RMS of the signals. Note that peak values from the entire trials period are well above the highest safety limits given in Table 1. Those limits were exceeded on many occasions during the trials with no reported injuries to boat crews or civilian test personnel. Also, the Max SSA values are more representative of boat motions prior to entering the ramp since the ramp impact accelerations were typically one or two instantaneous events within the two to ten minute data record. Table 10 Summary Small Boat Motions During Launch and Recovery Testing X Accel (g) Y Accel (g) Z Accel (g) Roll (deg) Pitch (deg) Max SSA All Data Max SSA Sea State Max SSA Sea State Peak All Data A more relevant presentation of small boat accelerations is given in Figure 7. This figure shows the number of peak accelerations in each direction exceeding the Safety Level limits given in Table 1. Note that this data includes all tests and is not representative of accelerations measured at optimum ship speed and heading. Still, these results are useful for analysis of the causes of limited operability and for selection of the safety level appropriate for these operations. Keeping in mind that Safety Level 1 is intended to protect personnel who are standing, originally on public transportation; it is evident from the figure that this safety level would result in very limited launch and recovery operability. Virtually all of the tests resulted in longitudinal accelerations exceeding Safety Level 1, indicating that landing on the ramp in any condition will cause this level of longitudinal acceleration. Although not as pronounced, this same observation applies to lateral accelerations. Assuming that Safety Level 2 or Safety Level 3 is appropriate for Coast Guard crews; results shown in Figure 7 indicate that lateral accelerations dominate operability. Using the previously defined categories, Human Factors are the main cause of unsafe launch and recovery operations. Paper No. Sheinberg 8

9 Number of Occurences X > Level 1 X > Level 2 X > Level 3 Y > Level 1 Y > Level 2 Category Figure 7 Distribution of Small Boat Peak Accelerations During Launch and Recovery Tests This term is somewhat misleading since lateral accelerations can be caused by stern sway motions and the handling characteristics of the specific boat, in addition to the skill of the operator. Nevertheless, alignment and lateral velocity of the small boat relative to the ship are the primary causes of accelerations that exceed Safety Level 2 or Safety Level 3. Another result related to small boat accelerations relevant for stern ramp design and specification of the appropriate safety level, is the short duration of the peak accelerations. In all cases where the largest accelerations were measured, rise times were very short and the peaks were very sharp. This is illustrated for a touch-and-go test in 5 ft significant head seas in Figure 8 where rise times for the X, Y and Z accelerations are 0.03, 0.02 and 0.02 seconds, respectively. This behavior is indicative of contact between two hard surfaces, and in fact the support rail material on the WMSL is hard with limited energy absorbing capability. In addition, there are knuckles in the rails when viewed from the top, and the small boat appeared to impact these knuckles during a majority of the recoveries. Y > Level 3 Z > Level 1 Z > Level 2 Z > Level 3 Based on experience gained during these trials, it would appear that Safety Level 3 is the appropriate criteria for Coast Guard crews. These acceleration levels were exceeded on many occasions during the trials, with acceleration peaks approaching two times the specified limits, and no injuries or damage were reported. These crews are seated during launch and recovery and the LRI has shock-absorbing seats with good hand holds and back support for all crew members. In addition, the short duration of the largest accelerations as discussed above, limits exposure to these peak values. However, it was clear when discussing results with the crews that some limits are necessary, since they did express concern over the accelerations for those tests with the largest peaks. Ship Motions The next most relevant criteria to be considered are the ship motions limits given in Tables 2 and 3. Based on the results from these trials, BERTHOLF appears to have very good seakeeping performance in the conditions tested. Table 11 provides a summary of maximum significant amplitudes and peak values for the relevant responses measured for all launch and recovery tests. This data set includes tests completed in significant wave heights between 2.5 and 8.4 ft. Table 11- Summary of Ship Motions During Launch and Recovery Testing Roll (deg) Max SSA All Data Peak All Data Y Accel Port (g) Pitch (deg) Z Accel Stbd (ft) Stern Heave (ft) Y Accel Stbd (ft) Stern Sway (ft) Sill Depth (ft) Z Accel Port (g) Stern Rel. Level (ft) Max SSA All Data Peak All Data Acceleration 52: : : : : : Time (min:sec) X Acceleration Y Acceleration Z Acceleration Figure 8 Small Boat Impact Acceleration Time Histories 5 ft Significant Head Seas With the exception of stern sway motion, all of the SSA values given in the table are less than the limits specified in Table 2 and Table 3. In fact, the peak values measured during the launch and recovery tests are less than the criteria, again with the exception of stern sway. It should be noted that no tests were completed in beam seas for the larger sea conditions, so roll angles given in the table do not include the beam sea conditions in seas larger than Sea State 3. The relatively large stern sway motions almost certainly contributed to the fact that lateral accelerations dominated the small boat motions. As shown in Table 9, a large number of tests were conducted in bow and stern quartering seas where stern sway motions would be most significant. Table 12 gives a similar set of statistics measured during the water-flow tests. This data set also includes water levels and velocities in the ramp. Statistics from this data Paper No. Sheinberg 9

10 set should have more significance that those given in Table 11 since measurements were taken over thirty-minute time periods at constant heading a speed, as compared to several minutes for the results given in Table 11. For the results given in Table 12, all maximum significant amplitudes, including stern sway, are within the limiting criteria given in Tables 2 and 3. This data set includes testing in significant wave heights ranging from 2.5 to 6.2 ft. the standpoint of ship pitch angle, beam seas would be the preferred launch and recovery heading, and head seas should be preferred over bow quartering seas. Again, the SMP predictions provide a good indication of ship response, except that trials data exhibited larger than predicted pitch in beam seas. This could also be due to the bi-directional nature of the encountered sea conditions. Table 12- Summary of Ship Motions During Water-flow Testing Statistic Max Significant Single Amplitude Peak Value Relative to Mean Statistic Sill Depth Center (ft) Sill Water Level Port (ft) Stern Stern Sway Heave (ft) (ft) Sill Water Level Stbd (ft) Stern Rel Water Level (ft) Sill Water Vel Fwd (ft/sec) Pitch Angle (deg) Sill Water Vel Aft (ft/sec) Roll Angle (deg) Pitch SSA/H1/3 (deg/ft) Sea State 3 Sea State 4 SMP Sea State 4 Max Significant Single Amplitude Peak Value Relative to Mean Rel. Swell Direction (deg) In addition to comparisons to operating criteria, measured ship motions can be analyzed to determine variations in the relative difficulty of launch and recovery operations with ship heading and speed. Although ship roll was not a factor during this test program it could be a significant factor if launch and recovery operations were conducted in beam seas. Figure 9 shows the variation in the roll quasi-transfer function (roll SSA / significant wave height) with ship relative heading for the water-flow tests. For comparison, predictions from the U.S. Navy s Ship Motions Program, SMP 95 are included on the figure. The fact that trials results show non-zero roll angles for head and following seas is probably due to the bi-directional nature of the sea state as discussed previously. Except for head and following seas, the SMP predictions are a good indication of the shape of the roll response curve, and reasonably predict peak responses. Figure 10 Pitch Quasi-Transfer Function versus Ship Relative Heading Figure 11 gives quasi-transfer functions for stern heave and stern relative water level versus heading for these same tests. In this case the trends are not well defined, and the relative water level is very scattered as compared to both absolute heave and pitch angle. For the range of wave heights and periods tested, trials data would suggest that head seas are always preferred. The SMP prediction for absolute heave shows general agreement with the magnitudes of measured absolute heave, and except for the following seas condition is in reasonable agreement in terms of shape. The SMP predictions indicate minimum stern heave in following seas, and that beam seas and head seas cause similar stern heave responses Roll SSA/H1/3 (deg/ft) Stern Heave SSA/H1/3 (ft/ft) Stern Relative Heave Stern Absolute Heave SMP Absolute Heave 0.20 Sea State Sea State 4 SMP Sea State Rel. Swell Direction (deg) Figure 9 Roll Quasi-Transfer Function versus Ship Relative Heading Figure 10 provides a similar presentation of pitch results from the water-flow tests. This figure indicates that from Rel. Swell Direction (deg) Figure 11 - Stern Heave Quasi-Transfer Function versus Ship Heading The large variations in relative heave could be in part due to the bi-directional nature of the seas or problems with the acoustic water level sensor used for these measurements. It Paper No. Sheinberg 10

11 should also be noted that the seakeeping program (VERES) used to develop the criteria for relative heave at the stern was based on strip-theory and could therefore not include the reflected or ship-generated waves in the relative heave predictions. Figure 12 gives the measured quasi-transfer function for stern sway and a comparison to SMP predictions. For this response, SMP under-predicts the magnitude across the complete range of headings. Although SMP was not used for the development of criteria in the previous study, a similar program was used, and this could be an explanation for the fact that stern sway is the only response measured during the trials that exceeded the previously developed criteria. Another possible explanation could be translation of measured sway from the center of gravity to the stern using yaw angle. These results were calculated by combining test data for similar wave heights, headings and speeds, and assume a threshold depth of 2.3 ft (mean sill depth) and a threshold time duration of 5 seconds (criteria given in Table 3). Sill Depth SSA/H1/3 (ft/ft) Rel. Swell Direction (deg) Sea State 3 Sea State 4 Stern Sway SSA /H1/3 (ft/ft) Sea State 3 Sea State 4 SMP Sea State 4 Figure 13 Sill Depth Quasi Transfer Function versus Ship Relative Heading Results given in Figure 14 indicate that ramp availability decreases by approximately 10% for an increase in significant wave height from 4.0 to 7.0 ft. Even at the higher of these two wave heights, conditions are suitable for ramp entry 80% of the time Rel. Swell Direction (deg) 110% 100% Figure 12 Stern Sway Quasi-Transfer Function versus Ship Relative Heading Another measure of the effects of ship motions on launch and recovery operations is the variation of sill depth at the ramp entrance due to a combination of ship motions, relative wave height outside the ramp and water-flow within the ramp. Figure 13 shows the quasi-transfer function for sill depth variations versus relative heading. Based solely on sill depth variations, these results also indicate that beam seas would be the best heading for launch and recovery and head seas are preferred over bow quartering seas. Ramp Availability Ramp availability is related to sill depth and is defined as the percentage of time where the sill depth exceeds a specified threshold value (typically the forward draft of the small boat) for a specified minimum time duration (believed to be necessary for ramp entry). In addition to ship stern motions and relative wave height, it is also influenced by the water-flow characteristics of the specific ramp design. Figures 14 through 16 give measured ramp availability percentages versus significant wave height, relative heading and ship speed from the launch and recovery tests. Availability 90% 80% 70% 60% 50% H 1/3 (ft) All Data Linear (All Data) Figure 14 Variation in Measured Ramp Availability with Wave Height Figure 15 show the variation in availability with ship relative heading by sea state. These results are similar to those given above in Figure 10 and 13 in that beam seas seem to provide the best ramp availability percentages. Variations in availability with ship speed are illustrated in Figure 16. Sea State 3 results included only two nominal speeds and there appears to be little difference in these two speeds in terms of ramp availability. Results given for Sea State 4 could possibly be interpreted to show an optimum speed range near 5 knots, but the individual data point variations are much larger than this trend. A more reasonable interpretation would be that availability is relatively constant for ship speeds up to 6 knots and then falls off slightly at higher speeds. Paper No. Sheinberg 11

12 110% where suitable conditions for entry are continuously maintained for this minimum time duration decreases. Availability 100% 90% 80% 70% 60% 50% Swell Relative Direction (deg) Sea State 3 Sea State 4 Poly. (Sea State 3) Poly. (Sea State 4) Figure 15 Variation in Measured Ramp Availability with Relative Heading A second use of the availability data is assessment of the effects of water depth required at the ramp sill. Figure 18 gives the variation in predicted availability with threshold water depth, and shows an almost linear decrease from 100% at 2.0 ft to 0% at 3.8 ft threshold depth. Again, these predictions indicate that as the required depth for entry increases, the amount of time suitable for entry decreases. Although these results cannot be directly interpreted as requirements for sill depth, they do indicate the importance of that design parameter. For the trials case with a sill depth of 2.3 ft, reasonable values for ramp availability result for threshold depths up to approximately 120% of the basic sill depth. It would seem reasonable that any decrease in the sill depth would result in a corresponding decrease in availability for a given boat draft. 110% 95% 100% 90% 90% 85% Availability 80% 70% 60% Sea State 4 Sea State 3 Poly. (Sea State 4) Linear (Sea State 3) Availability 80% 75% 70% 65% 60% SS 4, 30 deg, 5 knots 50% Vs (knots) 55% 50% Duration (sec) Figure 16 Variation in Measured Ramp Availability with Ship Speed The ramp availability results presented in Figures 14 through 16 are based on an assumed threshold time of 5 seconds and a threshold depth of 2.3 ft. If it is determined that a longer threshold time or a deeper threshold depth is required for successful boat entry into the ramp, the availability percentages, defined as the percent time when recoveries are possible, would decrease. During the trials it was noted that the coxswains were having difficulty maintaining the position and heading of the LRI while waiting for conditions in the ramp to become suitable for entry. Based on this observation, and the fact that the 5 second ramp availability criteria given in Table 3 was stated to be the minimum time, trials results were analyzed to assess how predicted availability would vary with threshold time. Results are shown in Figure 17. For an example case of Sea State 4 at a 30 degree relative heading and a ship speed of 5 knots, ramp availability would decrease from approximately 90% at 5 seconds to approximately 75% at 8 to 10 seconds. Again, these are predicted values of the percentage of time when successful ramp entry should be possible given different assumptions on the time required for successful entry. As the assumed minimum time for entry increases, the percentage of time Figure 17 Variation in Ramp Availability with Threshold Time Availability 120.0% 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% Minimum Depth (ft) SS 4, 30 deg, 5 knots Figure 18 Variation in Ramp Availability with Threshold Depth Ramp Water Flow Characteristics During launch and recovery operations, the boat ramp is subject to dynamic water flow. The hydrodynamics of the well are generally governed by the motion of the stern and the local wave field. Based on a frame of reference fixed to the stern of the vessel, the primary flow dynamic is an Paper No. Sheinberg 12

13 oscillation in the water elevation within the ramp at a period on the order of the wave encounter period. However, higher order flow effects develop as a result of the complex geometry of the ramp. These effects include sloshing, mixing, and draining. Previous work (Werenskoid, 2003) has demonstrated that proper management of water flow within the ramp can improve launch and recovery operations. This includes maintaining sufficient sill water elevation for the boat to enter the ramp and reducing the amount and velocity of flow up and down the ramp. Controlling the sill elevation improves the ramp availability, while minimizing flow up and down the ramp increases controllability as the boat enters the ramp. Two types of analyses were employed to gauge the effectiveness of the existing boat launch and recovery system installed on BERTHOLF. The first analysis addressed the issue of water flow up and down the ramp, and the second analysis quantified the contribution of the different modes of stern motion to the variations in ramp water levels. The time history of the water elevation in the false bottom was used to estimate the ratio of drain time to fill time for a given cycle. An algorithm was developed to search the time history for local extrema. Local minima corresponded to the times at which new cycles of filling began and the previous cycles of draining ended. This is illustrated in Figure 19 by way of example. The blue line represents the time history of the water elevation while the red lines correspond to the beginning of a fill cycle and the green lines correspond to the beginning of a drain cycle. The ratio of drain time to fill time was then determined by the cycle times for a subsequent pair of fill and drain cycles. Figure 19 Drain and Fill Cycles Results exhibit a great scatter in the ratio of drain time to fill time (D/F). However, clear trends appear when the ratio for a given occurrence is plotted according to the overall change in elevation during that occurrence. This delta is the difference in the water elevation from the points identified as the start and end of the fill cycle and is illustrated in Figure 19. The relationship appears to be linear, with the elevation delta increasing as the ratio increases. This relationship can clearly be seen in the Figures 20 and 21. It should be noted that because this analysis was limited to cases where boat operations were not conducted, it is not possible to correlate the results of this analysis with operability or impact acceleration levels. For the cases where the ramp was empty during the run (no boat in the ramp), the D/F ratio is typically larger than one. This result indicates that the upper portion of the ramp is managing the water flow more effectively than a solid ramp. Generally, a shallower slope in the relationship between the D/F ratio and the elevation change indicates that the ramp is more effectively reducing the mass flow rate of the draining water. Conversely, a steeper slope in the relationship means that more water is running back down the ramp. This is an indication that the false bottom is becoming fully saturated and the upper portion of the ramp is not effectively absorbing incoming water. The slope of a linear regression through these data sets varies by relative heading. The slope tends to be steepest in head and following seas, while the slope is the shallowest in beam seas. The smaller slopes in the beam sea conditions are likely a result of the reduced pitch motion. Thus, when water does manage to fill the false bottom it is allowed to drain without the influence of pitching. For the oblique heading conditions, the slope of the linear regression tends to fall somewhere between the head/following and beam conditions. From a water management standpoint, boat launch and recovery operations are best performed in beam or oblique seas. At these headings, the upper portion of the ramp drains at a relatively slower rate for a given change in elevation. The slower drain rate should correspond to less water running out the ramp as the boat attempts a recovery. The second type of analysis, spectral analysis, was performed only on the 22 motion and water flow tests. For this analysis, the time domain stern motion components and the sill water elevation were translated to the frequency domain using an FFT technique. The results of these transformations were spectral density distributions for the pitch-induced stern heave, the body heave at the longitudinal center of gravity, the total heave at the stern, and the sill elevation with respect to the mean sill level. Water in the ramp tends to be more responsive to excitation at higher frequencies (0.10 to 0.15 Hz). The water can be excited at lower frequencies (0.05 to 0.10 Hz); however, the response of the water level per unit heave motion is considerably smaller than at the higher excitation frequencies. These responses can be illustrated by way of spectral plots, two of which are shown in the Figures 22 and 23 for empty ramp conditions. In head seas, the stern motion and the sill elevation tend to be bi-modal spectra as seen in Figure 22. The body heave is the primary component of the total stern heave at the lower modal frequency while the pitch-induced heave is the primary component at the higher frequency. Thus, the Paper No. Sheinberg 13

14 Figure 20 - Run 283: D/F Trend with Elevation Change (Head Seas, 5 kts, H 1/3 = 6.2 ft) Figure 22 - Stern Motion & Sill Depth: Head Seas Figure 21 - Run 101: D/F Trend with Elevation Change (Quartering Seas, 5 kts, H 1/3 = 3.1 ft) pitch-induced motion tends to be the driving factor in the sill level response in head seas, as would be expected. The water level spectra for bow quartering conditions tend to be multi-modal, though lower in magnitude than head sea conditions. The pitch motion of the ship is considerably reduced in beam seas. Here, the body heave primarily drives the stern heave motion. However, the magnitude of the water elevation spectra is reduced. This indicates that in beam seas, the amount of activity in the ramp water level is less than all other conditions. In stern quartering headings, the water level spectra tend to be single-peaked at the higher excitation frequency. This can be seen in Figure 23. While a single-peaked response results in water level oscillations at a regular period, the shorter period of the oscillation may be less desirable to the small boat coxswain. The response of the water in the ramp tends to be single-peaked and occur at the lower excitation frequencies in following conditions. This can likely be attributed to the decrease in wave encounter frequency in following seas. The response per unit motion is relatively small at the lower frequencies, which is consistent with the general results of the analysis. Figure 23 - Stern Motion & Sill Depth: Quartering Seas The significant magnitude of the sill elevation spectra provides insight into the amount of water activity in the ramp. This significant height is based on the area underneath the spectral density distribution, and the derivation is similar to the calculation of a significant wave height. Comparing the significant height of the sill depth to the significant wave height produces a non-dimensional measure of the amount of water level activity in the stern relative to the local wave field. This ratio varies by relative heading and the relationship is shown in Figure 24. The ratio of significant sill elevation change to the significant wave height is analogous to a spectral transfer function. From head seas to beam seas, the significant sill depth is about half of the significant wave height. Beyond beam seas, the ratio increases to roughly 0.7 in stern quartering seas and 0.8 in following seas. Although few data exist at relative headings beyond 180 degrees, the transfer function appears to be symmetrical about that relative heading. According to this analysis, desirable headings for boat recovery include bow and beam seas. The amount of activity in the water level at the sill is relatively less than head and following sea conditions. Calmer water in the Paper No. Sheinberg 14

15 ramp will result in greater ramp availability and less water flow up and down the ramp. Boat Handling Characteristics As stated above, lateral accelerations dominated the cases where small boat accelerations exceeded any of the safety criteria. Although many of these cases can be attributed to the human operator, the handling characteristics of the small boat operating in the wave and wake environment behind the ship could be significant contributors to this operating the controls at varying levels as shown in Figures 26 and 27. Small Boat Impact Loads An area where design data is required not related to operability is ramp impact loads acting on the bottom of the small boat. These impact loads were measured using strain gages installed on bottom longitudinal stiffeners. To convert the strain measurements to impact pressures, a finite element model of the structure was developed and Figure 24 - Sill Depth Quasi-Transfer Function situation. During the trials, the small boat was instrumented to monitor operator control actions as well as position and heading relative to the ship. A typical time history of relative heading and position is shown in Figure 25. The arrows in the figure represent snapshots of the velocity magnitude and direction and the line is the resulting track of the LRI, both relative to the ship. This is a case in relatively mild Sea State 3 conditions. Note the multiple changes in direction of the relative velocity vector and the 15 ft variation in lateral position. Both of these are significant variations considering that the ramp is 2 ft wider than the boat. Figure 26 illustrates the change in LRI steering activity with significant wave height. The results are given as the significant single amplitude of steering commands as a percentage of maximum available steering travel. For example, a value of 20% would mean the coxswain is routinely moving the wheel plus and minus 20% of maximum travel. Figure 27 shows variations of this same parameter across relative heading. Clearly, for the conditions tested, the coxswains were working harder to maintain heading and alignment for bow quartering seas than for head seas or headings approaching beam seas. Trends were also investigated for steering activity variations with speed, and none were apparent. Variations in heading angle were also checked to determine if there were trends with wave height, relative heading or speed, and none were apparent. This would seem to indicate that the coxswains were successfully maintaining heading variations within a fairly constant range by Figure 25 LRI Position and Heading Relative to BERTHOLF Just Prior to Recovery Steering SSA (%) 60% 50% 40% 30% 20% 10% 0% Significant wave Height (ft) Figure 26 LRI Steering Activity versus Wave Height unit pressures were placed on each measurement area where gauge pairs were located. The responses at all gauge pairs were computed for a load on each gauge pair. The resulting matrix was inverted to get a data reduction matrix. Measured strains on all gauges were multiplied by the Paper No. Sheinberg 15