Modifying a Military-Grade Glider For Coastal Scientific Applications [Sea Technology]
(Sea Technology Via Acquire Media NewsEdge) Clider Initially Made for ONR Deployed in Academic, Research Missions
Gliders are well-known in the industry for viding long-duration sensing platforms. The original, or legacy, gliders (Slocum, Seaglider and Spray) were developed under an initial Office of Naval Research (ONR) grant in the late 1 990s and early 2000s, with a common size and operating envelope based on the initial criteria that they be two-man portable (i.e., lighter than 50 kilograms) and able to operate in the open ocean.
While these requirements are suitable for a range of applications, they impose unrealistic constraints that make legacy gliders difficult or impossible to use in some environments, particularly coastal areas with large density gradients and strong currents. For example, these gliders cannot adaptively ballast the unit for continuous operation from freshwater to seawater without manual intervention or continuously vary the speed over a wide range.
To bring these characteristics to gliders, Exocetus Development LLC (Anchorage, Alaska) has been modifying ANT LLC's (Anchorage) Littoral Glider, now called the Coastal Glider, since purchasing its assets, intellectual property and manufacturing technology in October. The ANT glider had been developed with ONR funding during the past six years.
The Coastal Glider is undergoing modifications, which includes the installation of a dedicated science computer and improved communications and survivability.
Rated to 200 meters depth, the Coastal Glider is capable of self-ballasting from fresh- to ocean water and has a variable speed capability to handle nearshore currents up to 2 knots. During initial design stages, potential users requested a maximum speed up to 5 knots, but the general consensus was that 2 knots would be sufficient.
Buoyancy Engine. The requirement for adaptive ballasting meant that the buoyancy engine (BE) volume be at least 3 percent of the vehicle volume. After the glider's hydrodynamic design was completed, it was determined that a 2 percent variation was needed. Thus, the BE volume of 5 liters represents 5 percent of the vehicle volume.
Several types of BE drive mechanisms were investigated. The simplest mechanism for driving the BE is a ballscrew arrangement. Ball screws, however, are subject to being back driven if there is no locking mechanism or locking current. These options would add excessively to the complexity or the power consumption, so the ball drive was eliminated. Alternatively, an Acme drive could have generated the necessary high drive forces without the back-drive issue, but its efficiency is very low. For these reasons, a hydraulic system was selected.
Hydro-Leduc's (Azerailles, France) Model PB32.5 microhydraulic pump was investigated, as it had been successfully used in profiling buoys and other gliders. The best efficiency point of this pump occurs at an operating pressure of 34.5 megapascals. This is significantly higher than the 2-megapascal water pressure at the maximum depth of 200 meters. A reverse-amplifier BE system was designed so that, at maximum depth, the BE pump operates at the best efficiency point. Coupling the pump to a high-efficiency DC motor resulted in a BE efficiency of 70 percent.
Pitch and Roll System. The pitch and roll components were selected such that a single mass (the battery) could be used for both pitch and roll control. This was accomplished by selecting a high-efficiency commercial off-the-shelf (COTS) linear actuator for the pitch system with an integral linear potentiometer for position feedback. This actuator was connected to an adaptor plate, which was in turn connected to the battery mass. The linear actuator is equipped with an integral slip clutch to prevent damage in case it is overdriven.
The battery mass was designed so the center of gravity of the battery was 2 centimeters below the glider's centerline. The layout of other components was designed so the glider's center of gravity rested on its centerline.
The roll actuator is also a COTS DC motor, which is mounted on the adaptor plate and, thus, moves with the battery mass. The roll system is connected to the battery mass by a drive chain, and measurement of the battery roll is affected by the use of a separate potentiometer engaged with the chain. Limit switches are also integrated into the adaptor plate to prevent damage if the roll system is overdriven.
All of the drive components (BE pump, BE valve, pitch motor and roll motor) are located in the front of the glider to keep the electrically noisy equipment away from the communications equipment, navigation equipment and science equipment in the aft electronics bay.
The arrangement of these components dictated a larger diameter than is found on other gliders. The outside diameter of the main hull measures 0.32 meters for compatibility with equipment for handling small-size standard torpedoes.
Battery and Power System. The glider runs on a lithium battery, with all electronics designed to be powered by 1 8 to 32 volts DC.
Power use in a glider is very irregular. The glider will go for long periods of time on the descent or ascent using very little power. Most of the motive power is used in a few seconds at the bottom inflection. This type of large draw is difficult for batteries to provide and tends to shorten battery life.
For this reason, an ultracapacitor pack was installed, from which all of the main actuators are powered. This also helps to decouple the motors from the more sensitive electronic equipment. The ultracapacitors are located below the pitch, roll and internal BE components.
Electronics Bay. The electronics bay comprises a dry cylindrical volume that has a 19.1 -centimeter inner diameter and is 30.5 centimeters long. Approximately 80 percent of this volume (7 liters) is available for sensor integration. The remaining section is reserved for glider communications and navigation equipment.
Communications and Navigation. The standard glider is equipped with three modes of communications: WiFi, Freewave and Iridium. The glider also has a GPS unit for determining location. All communications are facilitated by an antenna mounted on the aft centerline of the glider.
All three communications modalities use the same interface and command structure. Wi-Fi is generally used in the shop or on the deck to communicate with the glider for uploading mission files and downloading large data files, or for doing routine maintenance tasks. Freewave line-of-sight communications is used when launching and retrieving the glider. Iridium monitors and controls the glider when it is out of range of other communications.
The communications system allows for telemetry of engineering data via Iridium to either a server at Exocetus or a specially configured customer server. Data includes position information and glider status, battery voltage, average power consumption and any warnings or faults. Data from the science computer is collected and stored on board but is not available via Iridium. All communications electronics are contained in the lower portion of the electronics bay on a single circuit board, which also supports an inertial measurement unit (IMU), which is a MicroStrain (Williston, Vermont) 3 DM-GX, and the electronics for the glider's depth sensor.
The IMU provides heading, pitch and roll information to the navigation software and also provides Z-axis acceleration data to filter the depth-sensor signal. The existing IMU cannot be reliably used for inertial navigation due to high drift rates.
Initial Development and Testing
The process of determining the overall desirable specifications began in 2005 by ANT LLC, followed by building and testing a prototype in 2006. By 2007, a full version of the glider's vehicle control software was ready for initial testing. Three prototype units of the Littoral Glider (LG) were built and designated LG01 through LG03.
Initial navigation results were very poor, with the vehicle unable to hold a heading due to an incorrect IMU calibration procedure. A new calibration method and software utility was developed and implemented, resolving navigation issues.
Tests also showed the reliability of the initial design of the roll control system to be poor. Several modifications were tested before implementing the present design, which has not had problems with reliability or operation.
Another problem was the reliability of the internal communications bus between the main navigation processor and the electronics bay. The glider also exhibited quality control issues with leakage and damage in external cables, and with an antenna that resulted in lost communications modes. These issues have since been identified, corrected and tested.
In 2009, pilot production began, and 15 more LG units, LG04 through LGl 8, were built for ONR in the following two years.
The Coastal Gliders have logged 5,000 hours of combined flight time, and 12 of the initial 15 units are still in operation. The have been used in a variety of environmental monitoring applications.
ONR Involvement. ONR has directed two gliders to Georgia Tech and three to the Navy Postgraduate School. The gliders are being modified to incorporate acoustic sensors, specifically Wi Icoxon (Germantown, Maryland) vector sensor units. These organizations are conducting independent development of control-system modifications to implement application-specific behaviors.
ONR has also directed two gliders to Naval Undersea Warfare Center Division Newport. Equipped with AML Océanographie (Sidney, Canada) CTD units, one glider recently collected océanographie information for a NASA project, operating for three days at the mouth of a fjord near Tiniteqilaaq, Greenland.
During most of this testing, the glider was commanded at 1 .5 knots to compensate for the expected high currents (greater than 1 knot). Power consumption was correspondingly high at 1 4 watts average over the flight. Approximately 32 percent of the alkaline battery was used.
Korea Institute of Ocean Science and Technology. Two gliders were purchased by the Korea Institute of Ocean Science and Technology, formerly known as the Korea Ocean Research and Development Institute, in 2010 and 2011. Each was equipped with four omnidirectional acoustic sensors: one mounted on the nose cone, one on the top of the vertical stabilizer and one on each wing tip.
Two tests were conducted with LC1 9 in 201 1 off the east coast of Korea. The first test consisted of flying the glider 112 kilometers from the coast of mainland Korea to Ulung Island. Due to currents expected to be in excess of 1 knot, an average speed of 1 .5 knots was commanded.
As expected, the glider was carried north of the projected path and spent much of the time ferrying against the current, resulting in a speed over ground significantly less than the commanded speed. After six days, the glider arrived at the island, but recovery was delayed due to weather, which did not clear until after the battery was depleted and contact with the glider was lost.
Three months later, the glider was recovered in fishing net and returned undamaged to the institute. In a second outing, the glider encountered lower currents, with an average commanded speed of 1 .2 knots. The glider arrived at the island after two days and with substantial battery reserves.
The glider was then flown back to the mainland via a series of heading commands, without attempting to follow the waypoint tracks. The institute is planning additional tests.
University of Southern Mississippi. The University of Southern Mississippi has supplied a JFE Advantech Co. Ltd. (Kobe, Japan) RINKO Il dissolved oxygen sensor and a Wetlabs ECO FLNTU (Philomath, Oregon) combined turbidity and fluorescence sensor for integration onto a glider platform. The vehicle will study hypoxia in the Gulf of Mexico. The sensors have been integrated on LG18, and testing is scheduled to verify the ability of the glider to operate in the shallow gulf waters near the coast in November 2012.
Testing was scheduled for water depths from 10 to 30 meters. If the glider operates well, it will be flown into shallower water to determine minimum operational depths.
A program has been started in conjunction with the University of Alaska Fairbanks to modify the glider communications system to provide subsampled science data to the user via an Iridium RUDICS (Router-Based Unrestricted Digital Interworking Connectivity Solutions) pathway.
In addition, an emergency lift bag system, powered by CO2 cartridges, has been developed and was used on the glider during initial testing. This system was designed for temporary use and is not currently available to customers due to usability and reliability issues. A redesign to eliminate these shortcomings and include the emergency lift system as a standard system component is underway.
The authors would like to thank Dr. Mike Traweek from ONR for his continued support of the glider development. Original funding for the glider program was provided by ONR under contract N00014-07-C-0834. Followon funding was provided under contract NOOOl 4-1 1 -M-0069.
The authors would also like to thank David Trivett and James Martin of Georgia Tech, Dr. Kevin Smith and Tad Masek of the Navy Postgraduate School, Mr. Nathan Banks of the Naval Undersea Warfare Center, Dr. Yossup Park of the Korea Institute of Ocean Science and Technology, Dr.
Vernon Asper and Kevin Martin of the University of Southern Mississippi, and Dr. Peter Winsor of the University of Alaska Fairbanks for their support and willingness to share operational plans and information regarding the gliders.
For a list of references, contact Ray Mahr at firstname.lastname@example.org.
Ray Mahr. Ir. is the co-founder of Exocetos Devel- opment LLC. He has extensive experience in the design, development and sales of instrumentation and equipment systems for océanographie and un- derwater acoustics applications. As a former senior manager at three océanographie instrumentation companies, he directed implementation of new océanographie products and systems.
(c) 2012 Compass Publications, Inc.
[ Back To TMCnet.com's Homepage ]