Trident Scholar Abstracts 1999

Gas turbine engines are used commercially and in the military to produce large amounts of power for a small physical size. Applications of gas turbine engines include planes, ships, helicopters and generators. It is the engineers responsibility to pursue the most fuel efficient engine, while maintaining a high output power. In this continual race for improvement, increasing the efficiency of the gas turbine engine is vital to the national interests of the United States.

The work from the gas turbine engine is produced by the rotation of the turbine blades due to air flow. In a jet engine, the exhaust from the gas turbine engine results in thrust which pushes the aircraft forward. In order to produce the maximum power from the turbine, the turbine inlet temperature should be as high as possible. The inlet temperature to the turbine is limited by the material properties of the blades. One technique used to protect the blades is a "cooling film." This protection fails when the "cooling film" is washed away by strong vortices in the flow. As air flows through the turbine, a strong vortex (horseshoe vortex) is formed at the junction of the turbine blade and the hub. The horseshoe vortex combines with vortices from neighboring blades to form a complex flow pattern between the turbine blades. The vortices impinge along the turbine blade surface removing the "cooling film" from the blade, thus producing a "hot spot" beneath the horseshoe vortex. These "hot spots" push the material limits of the blades. The primary objective of this Trident Research Project is to modify the flow in this region by diverting the path of the vortex. This will allow for higher inlet temperatures in the turbine and thus increase the efficiency.

In this project, detailed measurements of the flow will be obtained to find the position of the vortex and the surface temperatures of the blades. Once this data has been mapped out, a wall jet will be installed along the turbine blades to produce a jet of air, pushing the vortex away from the hub regions of the blades. This diversion will prevent the vortex from removing the film cooling flow surrounding the blades.

This project represents a great opportunity to conduct valuable research and practical engineering design. The successful completion of this project will prove to be valuable in many Military and commercial applications. Experimental information gained from this project will further turbine efficiency research.

Heat pipes are hermetically sealed vessels that use the thermodynamic properties of a working fluid to transfer heat more efficiently than a single homogeneous material. The working fluid is evaporated on one end of the pipe and travels to the other end, where it condenses, releasing the large amount of heat that was required for the evaporation. The condensation then returns to the evaporator end via capillary action. If kept at a constant pressure, then the working fluid will evaporate and condense at a uniform temperature. In high temperature applications, it emits energy at a constant wavelength as well as being isothermal. This is particularly useful for thermophotovoltaic (TPV) energy conversion as photovoltaic cells are designed for optimal performance at certain energy wavelengths.

Flat heat pipes are anticipated to have increased power to volume ratios over conventional cylindrical designs due to their geometry as well as their potential for stacking or layering energy conversion modules with flat heat pipes. In addition to TPV applications, it is anticipated that flat heat pipes will also have many applications in cooling and other energy conversion devices.

The proper construction to withstand the pressures and vacuum inherent to the pipe's operation will be designed. In addition, methods of evacuating the vessel so proper amount of working fluid remains will be determined. The minimum operational thickness of the heat pipe that will allow for transport of working fluid will also be determined. While the principles of heat pipe operation seem simple, proper design of a heat pipe is rather complex.

By varying the position and application of the heat source as well as the orientation of the heat pipe, the heat pipe's true operating characteristics will be observed. Thermocouples, linked to a data collection system and mounted throughout the heat pipe will record the pertinent data. Data collected will be used design revision and analysis. Then the true potential of flat heat pipes will be seen.

As a Trident Scholar research project, it is proposed to extend an important work in support of the Cassini mission to Saturn. An extension to the current Saturnian magnetosphere can be developed by appending a realistic magnetotail which thus far the previous models have lacked. The magnetotail structure provides a mechanism for the rejoining of deflected solar wind particles after they have traveled past the planet following deflection from their original paths by interaction with the magnetospheric field. These particles generate a dawn to duskward cross-current as they re-merge to eventually comprise a neutral plasma streaming outwards (from the inner solar system). The cross-tail-current produces an additional magnetic field contribution which perturbs the entire planets model magnetospheric field. Through an extensive computer program, this cross-current can be modeled and its effects to the magnetosphere can be determined. The incorporation of a magnetotail to Saturn's magnetosphere will greatly increase the range of validity on the night side of the model magnetosphere as well as provide an important refinement to the day side. The cross-tail-current will first be incorporated into the Saturnian magnetospheric model produced by Maurice and Engle in 1995 (ME95) and then verified by comparing the results to the Voyager data. Once an equivalent or better fit is attained between the revised ME95 model and spacecraft data, the magnetotail will be incorporated into an extended ME95 model produced by Maurice, Engle, Blanc and Skubis produced in 1996 and the results will again be compared to the Voyager data. This will be helpful for the Cassini mission when it arrives in 2004 since the satellite will be spending a large fraction of its time at high Saturnian latitudes as Cassini executes its polar orbits.

The relationship of optics to physical processes occurring within the Chesapeake Bay will be established through an intensive study combining data collection and analysis with optical theory. Data collection will utilize a Conductivity-Temperature-Depth profiler for the acquisition of hydrographic data, a transmissometer for optical data, and sample collection for laboratory analysis of the suspended particulate load. Optical data collected in the fall and spring during times of peak physical variation can be analyzed to reveal the relation of optical properties inherent to the water column and the suspended particulate load. The suspended load can be expected to vary in composition in response to physical changes within the water column. A correlation between total beam attenuation, an optical property, and total suspended volume of the suspended particulate load will allow for the development of a robust predictive model of optical variability within the Bay based on understood seasonal physical variations. Radiative transfer theory can be utilized to establish the relation between the inherent optical properties of the water column and apparent optical properties. An understanding of this relation will allow an analysis of the ability of remote sensing instrumentation to forecast and nowcast physical estuarine processes through optical means. In addition, the geographic range of the predictive model can be analyzed through the application of data obtained from the Chesapeake Bay Observation System. Proper consideration and analysis of data will result in a robust prognostic model of both inherent and apparent optical variability in the Chesapeake Bay to identify and forecast the physical mechanisms at work in the estuary.

The intent of this project will be to design and build a methane combustion system whose control will be based upon a mathematical model of the combustion process. The proposed apparatus will incorporate a closed-loop control design that will automatically restrict the fuel-to-air ratio of the process to the values required for the most efficient level of combustion.

Existing methods use means of limited capacity that require individual tuning through a procedure of trial and error. The system that will be constructed will be free from the errors inherent within the operation of these designs, and will consequently be more capable of maintaining combustion at the desired level.

The system will operate by continuously measuring the quantities of carbon dioxide, carbon monoxide, and oxygen present within the exhaust of the combustion process. Based upon this analysis of the gases produced, a personal computer will be used to regulate the flow of methane to the burner. The PC will contain the required control algorithms, derived from a mathematical model of the combustion process, to accurately control the burner, and will permit easy modification of the control system when necessary. This combination of technologies, mathematical modeling and systems control design, will permit the reactions within the burner to be more tightly regulated.

The technology that will be developed within this project will allow industries dependent upon combustion processes to more rigidly control the progression and efficiency of these reactions. The final design will provide a system capable, despite fluctuations in load and varying environmental conditions, of limiting methane combustion to the narrow tolerances required for the most efficient level of the reaction.

In this project, various surface treatment methods will be used to roughen a metal substrate to allow for the development of an adequate mechanical bond with urethane. Development of such a mechanical bond may ultimately lead to the adoption of this process in the production of propellers utilized by the Navy.

The current propeller system production method used by the Navy is extremely difficult, time consuming, and expensive. Due to the increase in technology and noise reduction on the Seawolf and NSSN submarines, the cost and time required of producing these sophisticated systems has skyrocketed. Research is being performed by the Naval Surface Warfare Center in alternative propeller production methods.

In the new method the chemical bond between the urethane and the substrate weakens after exposure to the environment. Since the chemical bond is proving ineffective, this project will attempt to develop an adequate mechanical bond between the urethane and substrate. To do this, three evaluation tests will be performed; a 90 Peel (Bend) Test, a Fatigue Test, and the evaluation of surfaces under the Scanning Electron Microscope.

To treat the substrate surface, abrasive blasting techniques will be performed. Various pressures and exposure times will be used in both air blasting and water jet blasting of the surface. One type of blasting material will be selected to lower the number of variables that need to be evaluated. After treating, the surface will be evaluated under the Scanning Electron Microscope, coated by urethane, and finally evaluated by the Peel and Fatigue Tests. The bond strength must be at least 80 pounds per linear inch of width of peel strip when evaluated by the Peel test to be effective. The samples will then be placed in a salt water environment, under cathodic protection, simulating service conditions, and will be evaluated monthly to determine any degradations of bond strength.

If this project proves successful, the machining time required for current propeller systems will drop approximately 90 percent, the cost of production by roughly 50 percent, and the system might be utilized on later productions of the NSSN.

This project will involve the design and development of a non-invasive CH-46 helicopter aft gearbox fault detector. Non-intrusive detection of gearbox fault condition will save time and prove cost-effective in both manpower and materials. The detector will classify incipient faults based on vibration information taken from the aft gearbox, and it will be designed through a combination of digital signal processing and artificial neural network technology.

The vibration data, which was collected by Westland Helicopters Ltd, contains digitized information for eight different seeded fault conditions. This data will be manipulated through several methods of digital signal processing, such as the fast Fourier transform (FFT), chirp-z-transform, the short-time Fourier transform (STFT), and statistical analysis. Phase information will also be analyzed. Amplitude and frequency demodulation may also be performed in order to provide additional information about the vibration signal.

The information collected in the digital signal processing part of the project will be used to create a data vector needed in the classification phase of the project. In this part of the project, an artificial neural network will be designed and used to classify fault condition. A neural network is a system of interconnected units that are trained to compute a specific output as a non-linear function of their inputs. The neural network will be trained with information from the data vector, and will be tested with information not used for training the network.

The detector is expected to achieve nearly 100% classification of gearbox fault condition. It is also intended that the detector provide general classification so that it can be used to detect faults in different gearboxes that share similar components.

This Trident research project will focus on increasing the life of copper electrodes used in the resistance welding of aluminum work pieces employed in the automotive and aerospace industries. When the copper and aluminum make contact at the interface, there
is some undesirable mixing which results in the formation of a copper-aluminum eutectic phase. The welding temperature must be high enough to melt the aluminum, but the eutectic phase has a low melting temperature as compared to either pure copper or pure aluminum. It is believed that the eutectic melts away causing the copper to degrade more rapidly and the electrode lifetime to be shortened.

In order to solve the problem, an attempt will be made to alloy the copper with a very high melting temperature refractory metal such as molybdenum, niobium, and tungsten. Since the refractory metals are not readily soluble in copper, it will be necessary to use unconventional processing techniques to produce electrodes of uniform composition. These techniques include spray forming, rapid solidification, and ion implantation. Ideally, the new processing techniques will allow for a higher concentration and a uniform distribution of the refractory metals. This will be confirmed by metallurgical analysis involving optical and electron microscopy. Additionally the electrodes will be tested for both conductivity and welding lifetime.

The success of this research will be determined based on two criteria. The primary objective will be to increase the lifetime of the copper electrode as measured through weldability studies. A secondary objective will be to achieve the longer life at a reasonable cost so that it can be accepted as a new technique for producing electrodes for welding aluminum.

Advanced Composite Materials (ACM's) are becoming a viable alternative material for both structural and non-structural applications. The benefits of ACM's include higher strength-to-weight properties, excellent durability properties and good thermo-mechanical properties. The civil engineering community has recognized the merits of ACM's and is in the process of researching the use of these properties as mechanisms for controlling structural response. One such use is the rehabilitation of damaged concrete structures using a type of ACM known as Fiber Reinforced Polymers/Plastics (FRP) by attaching thin laminates to the tension face of the system.

The use of externally epoxy-bonded steel plates has been an accepted method of repair for years. The bonded plate serves as additional flexural reinforcement, and provides increased crack control for the concrete member. Replacing steel with FRP has many advantages including higher specific strengths and stiffness, resistance to acidic and alkaline corrosion, increased durability, lower cost and minimal maintenance.

This research is concerned with the ductile behavior of reinforced concrete beams externally retrofitted with fiber reinforced plastic (FRP) laminates. The objective of this work is to quantify the ductile response of the repaired beams as the anchorage length is varied. The anchorage length refers to the on-center length of the composite laminate attached to the concrete beam.

The objective will be achieved by first fabricating concrete beams that accurately model structural members, and fail predominantly in flexure. These beams will be stressed until they display initial cracking at which point, the thin composite laminates will be attached to the tension face using structural adhesives. The beams will then be cracked to ultimate failure and strain data collected. This data will be used to determine the ductility index of the system. A test matrix will use a feed back loop to optimize the anchorage length with respect to the system ductility.

This project will focus on the proposed NATO Multilateral Force (MLF) of the early 1960s. The MLF was a plan which called for the sharing of nuclear weapons within a NATO-governed, multinational fleet. The United States offered the MLF as an answer to demands by the other members of NATO to have a share in the alliance's nuclear planning and decision making. Despite receiving significant support, the MLF proposal failed, but its conception and appeal offer revealing insight into the United States' and NATO's nuclear weapons philosophy in the early 1960s as well as the cracks created in the alliance by nuclear weapons issues over that same time period.

In recent years, a large number of documents relating to the MLF have been declassified. These documents will be interpreted and analyzed in order to present a more recent and accurate history of the MLF. The project will focus on presenting the MLF from a naval perspective as well as a foreign policy perspective. It will be based on archival research done in both the United States and the United Kingdom, making it the first MLF study to be done from a multinational perspective. Furthermore, it will be the most significant study done on the issue since the end of the Cold War which should allow for an analysis that is free from the former East-West rhetoric. This project is intended to look at the new government documents available and determine how they affect and change previously held views on the entire MLF issue.

The project's purpose will be to develop an adaptive error-correction algorithm to improve digital communications in a noisy environment.

Much work has been done in communication electronics with the development of fixed error-correcting codes. By appending these codes to blocks of data during transmission, errors induced into this data stream by a "noisy" environment can be located and removed at the receiver. However, these fixed-length codes are designed to only remove errors up to a maximum amount. Therefore, they have to be designed longer and more complex in order to deal with worst case scenarios. In times when the environment is not "noisy", the bandwidth is wasted due to this over-coding, and the rate of data being transmitted is lowered. Using an adaptive algorithm will provide an improvement by maximizing the amount of data transmitted for any particular environment.

The first phase of the project will be to develop the algorithm on a computer with MATLAB and C++ programs. These programs will simulate coded data transmitted across a binary symmetric communications channel. The errors induced into the data will be counted upon their reception and subsequently corrected. From this error information, the adaptive algorithm will determine the necessary code length, and change according to the differing real time error rates. The second phase of the project will be to implement the system into hardware using digital signal processors, programmable logic and/or microprocessors.

In order to evaluate the effectiveness of the adaptive algorithm, the experimental results will be compared to the performance data of fixed codes and the gain produced by this implementation will be analyzed. More experimental work then will be performed to find the optimal implementation of this system and its theoretical limits.

Such an adaptable error-correcting coding scheme can be used where high speed and high reliability of transmitted digital data are both required in a communications system.

The goal of this project is to combine statistical methods with time-delay-embedding prediction methods on chaotic systems to yield an error analysis. A primary goal of the physical sciences has always been to predict the outcome of a system given its current conditions. Various scientific fields do make very accurate predictions on how systems evolve. Though, systems do exist where predictions are virtually impossible to make. Some of these systems involve random factors, others are completely deterministic. A chaotic system has the property that it produces data which appears to be randomly generated, but it is actually completely deterministic. Real-world data collected from such systems quickly diverge from models of even the simplest chaotic systems. Chaos theory gives a way to make predictions on chaotic data sets. The method known as time-series embedding analysis allows the evolution of a chaotic dynamical system to be predicted using only observed data from the system. The goal of this project will be to develop an error statistics theory to determine the quality of predictions made through the use of time-delay embedding.

For a prediction made through time-series embedding, confidence intervals will be determined, i.e. an interval which has a certain probability of containing the true prediction value. The project will develop and analyze the one-dimensional case as well as the more difficult multi-dimensional case. The theory will be tested on large sets of both computer-generated and experimentally-gathered chaotic data. With the large data sets, we will experimentally determine confidence intervals in which predictions should lie, and then compare them to the analytically determined intervals. While this project will involve many different data sets, the purpose of the project is not to analyze these specific data sets, but to develop a general analytical method which could theoretically be used on any chaotic system.