We have been intensely monitoring photometric variability in proto-planetary nebulae (PPNe) over the past 25 years and radial velocity variability over the past ten years. Pulsational variability has been obvious, in both the light and velocity, although the resulting curves are complex, with multiple periods and varying amplitudes. Observed periods range from 25 to 160 days, and the periods and amplitudes reveal evolutionary trends. We will present our observational results to date for approximately 30 PPNe, and discuss these results, including the search for period changes that might help constrain post-AGB evolutionary timescales.

NiTi (also known as Nitinol) is an almost equiatomic alloy of nickel and titanium and has many applications in various industries, such as biomedical, automotive, and aerospace. NiTi shape memory alloys undergo martensitic phase transformations under both thermal and mechanical loading and exhibit unique properties, such as superelasticity (SE) and shape memory effects (SME). Modeling the fatigue behavior of this alloy is very challenging due to the unique mechanical response of the material. Moreover, there are very limited studies on the fatigue behavior of this alloy under more realistic loading conditions, such as variable amplitude loading and multiaxial loading. In this study, strain-controlled cyclic experiments have been conducted in different conditions to study the variable amplitude fatigue behavior of superelastic NiTi. Nonzero mean strain/stress behavior of superelastic NiTi is investigated, and it is demonstrated that the classical fatigue models for mean strain/stress correction do not appropriately model the nonzero mean strain/stress fatigue behavior of superelastic NiTi. It is shown that, despite common metals (e.g., steel, aluminum, and titanium alloys), mean strain also affects the fatigue behavior of superelastic NiTi, as the resulting mean stress does not fully relax under cyclic load. Two energy-based fatigue models have been proposed based on the results in this study and provide acceptable correlation with experimental observations. The models proposed in this research, account for the effects of mean strain/stress and variations in cyclic deformation. The variations in the cyclic deformation can be due to several factors, such as slight changes in chemical composition, heat treatment processes, texture, etc. The predicted fatigue lives using the proposed fatigue model fall within scatter bands of 1.5 times the experimental life for constant amplitude loading. Analyses also show that the proposed total fatigue toughness parameter, ΣWt, together with the Rainflow cycle counting technique can accurately predict the fatigue life under more realistic loading condition, such as two-step (i.e. high-low and low-high) and variable amplitude load-paths.

In this study, a new consistent formulation coupling kinematics, thermodynamics, and kinetics with damage using an extended multiplicative decomposition of the deformation gradient that accounts for corrosion effects is presented. The technical approach used for modeling the corrosion behavior of magnesium alloys was divided into three primary steps. First, a predictive corrosion model was developed based on experimental corrosion observations. The experimentally-observed corrosion mechanisms of pitting, intergranular, and general corrosion on the AZ31 magnesium alloy were quantified in 3.5 wt.% NaCl immersion and salt spray environments using optical microscopy and laser profilometry to document the changes in the pit characteristics. Although both environments showed similar trends, the immersion environment was more deleterious with respect to intergranular and general corrosion. On the other hand, the salt-spray environment allowed deeper pits to form throughout the entirety of the experiments, which led to a substantial thickness drop (general corrosion) compared with the immersion environment. Next, the complete corrosion model based upon the internal state variable theory was formulated to capture the effects of pit nucleation, pit growth, pit coalescence, and general corrosion. Different rate equations were given for each mechanism. Following the formulation of the model, the aforementioned experimental work and experimental work on four other magnesium alloys (AZ61, AM30, AM60, and AE44), was used to validate the model.

In recent years the increasing demand for fuel efficient and less pollutant vehicles has stimulated the development of hybrid and electric vehicles. These vehicle platforms often incorporate drivetrains which utilize multiple power sources for vehicle propulsion in an effort to increase fuel mileage and reduce emissions. Coupling multiple power sources, such as an internal combustion engine and electric motor(s), has new challenges in drivetrain design. Understanding the torque and rpm relationships within the power transmission device used to combine power sources is fundamental to overcoming the design challenges associated with hybrid and electric vehicle platforms. Results from this research include the fundamental torque and rpm relationships that exist in a multiple-input, single-output power transmission device. These results were deduced from a test that incorporated two separate power inputs into a differential which combined to produce a single output. Testing displayed that a differential has the ability to function as an infinitely variable transmission (IVT). Additionally, the challenges associated with using a differential as a multiple-input, single-output device were identified. Recommendations for overcoming these challenges are also presented herein. This work provides the basis for future work in powertrain optimization for multiple-input, single-output transmission devices.

No description available.

No description available.

Experiments were carried out to investigate the effect of ambient air temperatures on the heat pump performance using a variable speed compressor. Ambient air temperatures were varied from 40 to 60 °F to simulate different seasons. The compressor frequencies of 45 Hz, 50 Hz, 55 Hz, and 60 Hz were studied to determine the optimal frequency under various heating loads. The investigation was carried out by showing the compressor power input, heating output, and coefficient of performance for each case. Thermal cycle analysis along with the heat exchanger theory was used to analyze the system energy balance, heat transfer rates, p-h diagrams, and coefficient of performance. The overall heat transfer coefficients were also determined for both the evaporator and the condenser. Only the capillary tube was used to regulate the refrigerant flow rate. The variable speed compressor system used in this study will help save energy when compared with the traditional steady speed system. The variable speed compressor system will hopefully provide a more comfortable and steady indoor temperature than the traditional system, which is controlled by only an on-off switch. The speed controlled compressor system proposed we believe will help saving more energy than traditional steady speed system. The variable speed compressor system will hopefully provide a more comfort and steady indoor temperature than the traditional system which is controlled by one switch. It is believed that the variable speed compressor system may allow the indoor temperature air to be steady-going and prevent the switch working frequently. / Mechanical Engineering

Artificial neural network (ANNs) is an information processing paradigm inspired by the human brain. ANNs have been used in numerous applications to provide complex nonlinear input-output mappings. They have the ability to adapt and learn from observed data. The training of neural networks is an important area of research and consideration. Training techniques have to provide high accuracy, fast speed of convergence, and avoid premature convergence to local minima. In this thesis, a novel training method is proposed. This method is based on the relatively new Smooth Variable Structure filter (SVSF) and is formulated for feedforward multilayer perceptron training. The SVSF is a state and parameter estimation that is based on the Sliding Mode Concept and works in a predictor-corrector fashion. The SVSF applies a discontinuous corrective term to estimate state and parameters. Its advantages include guaranteed stability, robustness, and fast speed of convergence. The proposed training technique is applied to three real-world benchmark problems and to a fault detection application in a Ford diesel engine. SVSF-based training technique shows an excellent generalization capability and a fast speed of convergence. / Artificial neural network (ANNs) is an information processing paradigm inspired by the human brain. ANNs have been used in numerous applications to provide complex nonlinear input-output mappings. They have the ability to adapt and learn from observed data. The training of neural networks is an important area of research and consideration. Training techniques have to provide high accuracy, fast speed of convergence, and avoid premature convergence to local minima. In this thesis, a novel training method is proposed. This method is based on the relatively new Smooth Variable Structure filter (SVSF) and is formulated for feedforward multilayer perceptron training. The SVSF is a state and parameter estimation that is based on the Sliding Mode Concept and works in a predictor-corrector fashion. The SVSF applies a discontinuous corrective term to estimate state and parameters. Its advantages include guaranteed stability, robustness, and fast speed of convergence. The proposed training technique is applied to three real-world benchmark problems and to a fault detection application in a Ford diesel engine. SVSF-based training technique shows an excellent generalization capability and a fast speed of convergence. / Thesis / Master of Applied Science (MASc)

Thesis. 1976. B.S.--Massachusetts Institute of Technology. Dept. of Earth and Planetary Sciences. / Microfiche copy available in Archives and Science. / Bibliography: leaves 34-35. / B.S.

Ensuring inspection performance is not a trivial design problem, because inspection is a complex and difficult task that tends to be error-prone, whether performed by human or by automated machines. Due to economical or technological reasons, human inspectors are responsible for inspection functions in many cases. Humans, however, are rarely perfect. A system of manual inspection was found to be approximately 80-90% effective, thus allowing non-confirming parts to be processed (Harris & Chaney, 1969; Drury, 1975). As the attributes of interest or the variety of products increases, the complexity of an inspection task increases. The inspection system becomes less effective because of the sensory and cognitive limitations of human inspectors. Any means that can support or aid the human inspectors is necessary to compensate for inspection difficulty. Augmented reality offers a new approach in designing an inspection system as a means to augment the cognitive capability of inspectors. To realize the potential benefits of AR, however the design of AR-aided inspection requires a through understanding of the inspection process as well as AR technology. The cognitive demands of inspection and the capabilities of AR to aid inspectors need to be evaluated to decide when and how to use AR for a dimensional inspection. The objectives of this study are to improve the performance of a dimensional inspection task by using AR and to develop guidelines in designing an AR-aided inspection system. The performance of four inspection methods (i.e., manual, 2D-aided, 3D-aided, and AR-aided inspections) was compared in terms of inspection time and measurement accuracy. The results suggest that AR might be an effective tool that reduces inspection time. However, the measuring accuracy was basically the same across all inspection methods. The questionnaire results showed that the AR and 3D-aided inspection conditions are preferred over the manual and 2D-aided inspection. Based on the results, four design guidelines were formed in using AR technology for a dimensional inspection. / Ph. D.