Mechanical & Energy Engineering Department Theses and Dissertations

Permanent URI for this collection

https://hdl.handle.net/1805/2088

Information about the Purdue School of Engineering and Technology Graduate Degree Programs available at IUPUI can be found at: http://www.engr.iupui.edu/academics.shtml

Browse

Now showing 1 - 10 of 179

Enhancing Mechanical Engineering Education Through a Virtual Instructor in an Ai-Driven Virtual Reality Fatigue Test Lab
(2023-08) Yahyaeian, Amir Abbas; Jones, Alan; Zhang, Jing; Du, Xiaoping
This thesis demonstrates the combination of virtual reality (VR) and artificial intelligence (AI) specifically exploring the practical application of Natural Language Processing (NLP) and GPT-based models in educational VR laboratories. The objective is to design a comprehensive learning environment where users can independently engage in laboratory experiments, deriving similar educational outcomes as they would from a traditional, physical laboratory setup, particularly within the realms of Science, Technology, Engineering, and Mathematics (STEM) disciplines. Using machine learning techniques and authentic virtual reality simulating educational experiments, we propose an advanced learning platform—Virtual Reality Instructional Laboratory Environment (VRILE). A key feature of the VRILE is an AI-powered instructor capable of not only guiding the learners through the tasks but also responding intelligently to their actions in real-time. The AI constituent of the VRILE uses the GPT-2 model for text generation in the field of Natural Language Processing (NLP). To ensure the generated instructions were contextually relevant and meaningful to lab participants, the model was trained on a dataset derived from an augmentation over user interactions within the VR environment. By pushing the boundaries of how AI can be utilized in educational VR environments, this research paves the way for broader adoption across other domains of engineering education. Furthermore, it provides a solid foundation for future research in this interdisciplinary field. It marks a significant stride in the integration of technology and education, encouraging more ventures into this promising frontier.
Modeling Acute Care Utilization for Insomnia Patients
(2023-08) Zhu, Zitong; Fang, Shiaofen; Ben Miled, Zina; Xia, Yuni; Zheng, Jiangyu
Machine learning (ML) models can help improve health care services. However, they need to be practical to gain wide adoption. A methodology is proposed in this study to evaluate the utility of different data modalities and cohort segmentation strategies when designing these models. The methodology is used to compare models that predict emergency department (ED) and inpatient hospital (IH) visits. The data modalities include socio-demographics, diagnosis and medications and cohort segmentation is based on age group and disease severity. The proposed methodology is applied to models developed using a cohort of insomnia patients and a cohort of general non- insomnia patients under different data modalities and segmentation strategies. All models are evaluated using the traditional intra-cohort testing. In addition, to establish the need for disease- specific segmentation, transfer testing is recommended where the same insomnia test patients used for intra-cohort testing are submitted to the general-patient model. The results indicate that using both diagnosis and medications as a source of data does not generally improve model performance and may increase its overhead. For insomnia patients, the best ED and IH models using both data modalities or either one of the modalities achieved an area under the receiver operating curve (AUC) of 0.71 and 78, respectively. Our results also show that an insomnia-specific model is not necessary when predicting future ED visits but may have merit when predicting IH visits. As such, we recommend the evaluation of disease-specific models using transfer testing. Based on these initial findings, a language model was pretrained using diagnosis codes. This model can be used for the prediction of future ED and IH visits for insomnia and non-insomnia patients.
Optimization and Characterization of Metal Oxide Nanosensors for the Analysis of Volatile Organic Compound Profiles in Breath Samples
(2023-08) Maciel Gutierrez, Mariana; Agarwal, Mangilal; Dalir, Hamid; Nalim, Mohamed Razi
Volatile organic compounds (VOCs) are byproducts of metabolic processes that can be uniquely dysregulated by various medical conditions and are expressed in biological samples. Therefore, VOCs expressed in breath, urine and other sample types may be utilized for noninvasive, rapid, and accurate diagnostics in a point-of-care setting. Currently, the most common methods for VOC detection include gas chromatography-mass spectrometry (GC-MS) and electronic noses (E-noses) that integrate nanosensors. Both methods present important advantages and challenges that allow their implementation for different applications. While GC-MS can be used to directly identify VOCs in complex matrices, it is a non-portable and high-cost instrument. On the other hand, E-noses are portable and user-friendly VOC detectors, but they do not allow for direct VOC identification or quantification. Among different VOC rich sample types, breath offers the advantage of being a virtually limitless source of endogenous biomarkers that can be implemented for noninvasive VOC detection. The presented thesis focuses on the optimization of the operating parameters (heater and sensor voltages) of a metal oxide (MOX) sensor and breath sampling techniques (sensor casing, breath fractionation, and exhalation volume) for their implementation in exhaled VOC analysis. In parallel, an in-house feature extraction algorithm was developed and implemented for the optimization of a MOX sensor composed of a tin oxide (SnO2) sensing layer. The optimized sensor parameters (heater voltage equal to 2 V and sensor voltage equal to 0.8 V) and breath sampling protocol (24 L of whole breath analyzed using the in-house sensor casing design) were tested with exhaled breath samples from distinct volunteers which could be successfully separated with 100% accuracy. The sensor response also showed a high degree of intrasubject reproducibility (RSD < 6%). Additionally, the sensor performance was further validated under ambient conditions, and sensor degradation was studied over the course of 3 months. Finally, sensor response to synthetic VOC profiles and individual VOC standards was explored. Optimized SnO2 sensors distinguished between VOC mixtures regardless of variations in relative humidity (RH) levels. Furthermore, the characteristic sensor response to VOC standards indicates that the sensors are most sensitive toward isopropanol by a factor of 1.15 in 45% RH and a factor of 3.58 in 85% RH relative to isoprene. To translate the potential of MOX sensors to point-of-care biomedical applications, there first exists the need to establish a reference of sensor baseline signals corresponding to exhaled breath samples from healthy individuals. SnO2 sensors and breath sampling methods were implemented for the collection of individual samples from 109 relatively healthy volunteers. 10 of these volunteers provided 9 additional samples over the course of six months. In parallel, exhaled breath samples were also analyzed by GC-MS to comprehensively profile VOCs present in the samples. The results from these experiments not only aid in the identification of the healthy breath signal baseline but also allow the exploration of VOC reproducibility over time. High variation between samples from distinct volunteers was observed, but samples longitudinally collected across volunteers could not be distinguished, alluding to the existence of a universal range of sensor signals that could describe the composition of exhaled breath from healthy subjects. Finally, results were compared with relevant confounding variables to better understand how VOCs are impacted by an array of factors that are not directly correlated to disease diagnosis. Sensor signals were significantly elevated in breath samples from male volunteers compared to samples from female subjects (p-value = 0.044). Interestingly, isoprene signals resulting from the GC-MS analysis were also higher in male subjects relative to females. No other relationships were identified between sensor signals and the confounding variables of interest. Future work would require a deeper understanding of sensor degradation and life cycle, along with sensor testing using a broader range of individual VOC standards and more complex VOC profiles. Additionally, further comparison between sensor signal and GC-MS signal of relevant VOC biomarkers present in breath would be beneficial. Nonetheless, the presented be leveraged in future investigations aiming to identify biomarkers for different medical conditions. Finally, the findings disclosed in the deposited thesis suggest the ability of a SnO2 nanosensor array to be implemented for breath analysis, providing a noninvasive, easy to use, and reliable diagnostic device in a point-of-care setting.
Research and Development of Electric Micro-Bus Vehicle Chassis
(2022-12) Coovert, Benjamin; Tovar, Andres; Nematollahi, Khosrow; El-Mounayri, Hazim
In this project, a chassis concept has been developed for a small electric vehicle ’minibus’. The vehicle is intended to be used as a transport between agricultural locations in Ethiopia to cities where the products can be sold. The objective is to develop a chassis that can house several different modular structures for the purposes of transporting refrigerated goods, a mobile power grid, or people. Literature studies have been conducted on current electric vehicle markets, battery markets, chassis materials, and optimal cross-sections. The battery housings have also been analyzed from an environmental perspective to account for conditions in Ethiopia. Based on this, it was found that a four-wheeled ’minibus’ design including space for approximately fourteen custom batteries is optimal. It is essential to keep in mind that this project has been carried out both on a conceptual level within the framework of a degree project as well as a production project for use in Ethiopian rural areas. This master thesis project aims to provide a solid benchmark for further development and research within the subject.
Fundamental Investigation of Direct Recycling Using Chemically Delithiated Cathode
(2022-12) Bhuyan, Md Sajibul Alam; Shin, Hosop; Zhu, Likun; Wei, Xiaoliang
Recycling valuable cathode material from end-of-life (EOL) Li-ion batteries (LIBs) is essential to preserve raw material depletion and environmental sustainability. Direct recycling reclaims the cathode material without jeopardizing its original functional structures and maximizing return values from spent LIBs compared to other regeneration processes. This work employed two chemically delithiated lithium cobalt oxide (LCO) cathodes at different states of health (SOH), which are analogous to the spent cathodes but free of any impurities, to investigate the effectiveness of cathode regeneration. The material and electrochemical properties of both delithiated SOHs were systematically examined and compared to pristine LCO cathode. Further, those model materials were regenerated by a hydrothermal-based approach. The direct cathode regeneration of both low and high SOH cathode samples restored their reversible capacity and cycle performance comparable to pristine LCO cathode. However, the inferior performance observed in higher current density (2C) rate was not comparable to pristine LCO. In addition, the higher resistance of regenerated cathodes is attributed to lower high-rate performance, which was pointed out as the key challenge of the cathode recycling process. This study provides valuable knowledge about the effectiveness of cathode regeneration by investigating how the disordered, lithium-deficient cathode at different SOH from spent EOL batteries are rejuvenated without changing any material and electrochemical functional properties.
Calibration and Validation of a High-Fidelity Discrete Element Method (DEM) based Soil Model using Physical Terramechanical Experiments
(2022-08) Ghike, Omkar Ravindra; El-Mounayri, Hazim; Tovar, Andres; Zhang, Jing
A procedure for calibrating a discrete element (DE) computational soil model for various moisture contents using a conventional Asperity-Spring friction modeling technique is presented in this thesis. The procedure is based on the outcomes of two physical soil experiments: (1) Compression and (2) unconfined shear strength at various levels of normal stress and normal pre-stress. The Compression test is used to calibrate the DE soil plastic strain and elastic strain as a function of Compressive stress. To calibrate the DE inter-particle friction coefficient and adhesion stress as a function of soil plastic strain, the unconfined shear test is used. This thesis describes the experimental test devices and test procedures used to perform the physical terramechanical experiments. The calibration procedure for the DE soil model is demonstrated in this thesis using two types of soil: sand-silt (2NS Sand) and silt-clay(Fine Grain Soil) over 5 different moisture contents: 0%, 4%, 8%, 12%, and 16%. The DE based models response are then validated by comparing them to experimental pressure-sinkage results for circular disks and cones for those two types of soil over 5 different moisture contents. The Mean Absolute Percentage Error (MAPE) during the compression calibration was 26.9% whereas during the unconfined shear calibration, the MAPE was calculated to be 11.38%. Hence, the overall MAPE was calculated to be 19.34% for the entire calibration phase.
Fabrication and Characterization of Lithium-ion Battery Electrode Filaments Used for Fused Deposition Modeling 3D Printing
(2022-08) Kindomba, Eli; Zhang, Jing; Zhu, Likun; Schubert, Peter
Lithium-Ion Batteries (Li-ion batteries or LIBs) have been extensively used in a wide variety of industrial applications and consumer electronics. Additive Manufacturing (AM) or 3D printing (3DP) techniques have evolved to allow the fabrication of complex structures of various compositions in a wide range of applications. The objective of the thesis is to investigate the application of 3DP to fabricate a LIB, using a modified process from the literature [1]. The ultimate goal is to improve the electrochemical performances of LIBs while maintaining design flexibility with a 3D printed 3D architecture. In this research, both the cathode and anode in the form of specifically formulated slurry were extruded into filaments using a high-temperature pellet-based extruder. Specifically, filament composites made of graphite and Polylactic Acid (PLA) were fabricated and tested to produce anodes. Investigations on two other types of PLA-based filament composites respectively made of Lithium Manganese Oxide (LMO) and Lithium Nickel Manganese Cobalt Oxide (NMC) were also conducted to produce cathodes. Several filaments with various materials ratios were formulated in order to optimize printability and battery capacities. Finally, flat battery electrode disks similar to conventional electrodes were fabricated using the fused deposition modeling (FDM) process and assembled in half-cells and full cells. Finally, the electrochemical properties of half cells and full cells were characterized. Additionally, in parallel to the experiment, a 1-D finite element (FE) model was developed to understand the electrochemical performance of the anode half-cells made of graphite. Moreover, a simplified machine learning (ML) model through the Gaussian Process Regression was used to predict the voltage of a certain half-cell based on input parameters such as charge and discharge capacity. The results of this research showed that 3D printing technology is capable to fabricate LIBs. For the 3D printed LIB, cells have improved electrochemical properties by increasing the material content of active materials (i.e., graphite, LMO, and NMC) within the PLA matrix, along with incorporating a plasticizer material. The FE model of graphite anode showed a similar trend of discharge curve as the experiment. Finally, the ML model demonstrated a reasonably good prediction of charge and discharge voltages.
Deep Learning Based Crop Row Detection
(2022-05) Doha, Rashed Mohammad; Anwar, Sohel; Al Hasan, Mohammad; Li, Lingxi
Detecting crop rows from video frames in real time is a fundamental challenge in the field of precision agriculture. Deep learning based semantic segmentation method, namely U-net, although successful in many tasks related to precision agriculture, performs poorly for solving this task. The reasons include paucity of large scale labeled datasets in this domain, diversity in crops, and the diversity of appearance of the same crops at various stages of their growth. In this work, we discuss the development of a practical real-life crop row detection system in collaboration with an agricultural sprayer company. Our proposed method takes the output of semantic segmentation using U-net, and then apply a clustering based probabilistic temporal calibration which can adapt to different fields and crops without the need for retraining the network. Experimental results validate that our method can be used for both refining the results of the U-net to reduce errors and also for frame interpolation of the input video stream. Upon the availability of more labeled data, we switched our approach from a semi-supervised model to a fully supervised end-to-end crop row detection model using a Feature Pyramid Network or FPN. Central to the FPN is a pyramid pooling module that extracts features from the input image at multiple resolutions. This results in the network’s ability to use both local and global features in classifying pixels to be crop rows. After training the FPN on the labeled dataset, our method obtained a mean IoU or Jaccard Index score of over 70% as reported on the test set. We trained our method on only a subset of the corn dataset and tested its performance on multiple variations of weed pressure and crop growth stages to verify that the performance does translate over the variations and is consistent across the entire dataset.
Physics-Based Modelling and Simulation Framework for Multi-Objective Optimization of Lithium-Ion Cells in Electric Vehicle Applications
(2022-05) Gaonkar, Ashwin; El-Mounayri, Hazim; Tovar, Andres; Zhu, Likun; Shin, Hosop
In the last years, lithium-ion batteries (LIBs) have become the most important energy storage system for consumer electronics, electric vehicles, and smart grids. The development of lithium-ion batteries (LIBs) based on current practice allows an energy density increase estimated at 10% per year. However, the required power for portable electronic devices is predicted to increase at a much faster rate, namely 20% per year. Similarly, the global electric vehicle battery capacity is expected to increase from around 170 GWh per year today to 1.5 TWh per year in 2030--this is an increase of 125% per year. Without a breakthrough in battery design technology, it will be difficult to keep up with the increasing energy demand. To that end, a design methodology to accelerate the LIB development is needed. This can be achieved through the integration of electro-chemical numerical simulations and machine learning algorithms. To help this cause, this study develops a design methodology and framework using Simcenter Battery Design Studio® (BDS) and Bayesian optimization for design and optimization of cylindrical cell type 18650. The materials of the cathode are Nickel-Cobalt-Aluminum (NCA)/Nickel-Manganese-Cobalt-Aluminum (NMCA), anode is graphite, and electrolyte is Lithium hexafluorophosphate (LiPF6). Bayesian optimization has emerged as a powerful gradient-free optimization methodology to solve optimization problems that involve the evaluation of expensive black-box functions. The black-box functions are simulations of the cyclic performance test in Simcenter Battery Design Studio. The physics model used for this study is based on full system model described by Fuller and Newman. It uses Butler-Volmer Equation for ion-transportation across an interface and solvent diffusion model (Ploehn Model) for Aging of Lithium-Ion Battery Cells. The BDS model considers effects of SEI, cell electrode and microstructure dimensions, and charge-discharge rates to simulate battery degradation. Two objectives are optimized: maximization of the specific energy and minimization of the capacity fade. We perform global sensitivity analysis and see that thickness and porosity of the coating of the LIB electrodes that affect the objective functions the most. As such the design variables selected for this study are thickness and porosity of the electrodes. The thickness is restricted to vary from 22microns to 240microns and the porosity varies from 0.22 to 0.54. Two case studies are carried out using the above-mentioned objective functions and parameters. In the first study, cycling tests of 18650 NCA cathode Li-ion cells are simulated. The cells are charged and discharged using a constant 0.2C rate for 500 cycles. In the second case study a cathode active material more relevant to the electric vehicle industry, Nickel-Manganese-Cobalt-Aluminum (NMCA), is used. Here, the cells are cycled for 5 different charge-discharge scenarios to replicate charge-discharge scenario that an EVs battery module experiences. The results show that the design and optimization methodology can identify cells to satisfy the design objective that extend and improve the pareto front outside the original sampling plan for several practical charge-discharge scenarios which maximize energy density and minimize capacity fade.
Modeling of Steel Laser Cutting Process Using Finite Element, Machine Learning, and Kinetic Monte Carlo Methods
(2022-05) Stangeland, Dillon; Zhang, Jing; Jones, Alan; Daehyun Koo, Dan
Laser cutting is a manufacturing technology that uses a focused laser beam to melt, burn and vaporize materials, resulting in a high-quality cut edge. Although previous efforts are primarily based on a trial-and-error approach, there is insufficient understanding of the laser cutting process, thus hindering further development of the technology. Therefore, the motivation of this thesis is to address this research need by developing a series of models to understand the thermal and microstructure evolution in the process. The goal of the thesis is to design a tool for optimizing the steel laser cutting process through a modeling approach. The goal will be achieved through three interrelated objectives: (1) understand the thermal field in the laser cutting process of ASTM A36 steel using the finite element (FE) method coupled with the user-defined Moving Heat Source package; (2) apply machine learning method to predict heat-affected zone (HAZ) and kerf, the key features in the laser cutting process; and (3) employ kinetic Monte Carlo (kMC) simulation to simulate the resultant microstructures in the laser cutting process. Specifically, in the finite element model, a laser beam was applied to the model with the parameters of the laser’s power, cut speed, and focal diameter being tested. After receiving results generated by the finite element model, they were then used by two machine learning algorithms to predict the HAZ distance and kerf width that is produced due to the laser cutting process. The two machine learning algorithms tested were a neural network and a support vector machine. Finally, the thermal field was imported into the kMC model as the boundary conditions to predict grain evolution’s in the metals. The results of the research showed that by increasing the focal diameter of a laser, the kerf width can be decreased and the HAZ distance experienced a large decrease. Additionally, a pulse-like pattern was observed in the kerf width through modeling and can be minimized into more of a uniform cut through the increase of the focal diameter. By increasing the power of a laser, the HAZ distance, kerf width, and region of the material above its original temperature increase. Additionally, through the increase of the cut speed, the HAZ distance, kerf width, kerf pulse-like pattern, and region of the material above its original temperature decrease. Through the incorporation of machine learning algorithms, it was found that they can be used to effectively predict the HAZ distance to a certain degree. The Neural Network and Support Vector Machine models both show that the experimental HAZ distance data lines up with the results derived from ANSYS. The Gaussian Process Regression HAZ model shows that the algorithm is not powerful enough to create an accurate prediction. Additionally, all of the kerf width models show that the experimental data is being overfit by the ANSYS results. As such, the kerf width results from ANSYS need additional validation to prove their accuracy. Using the kMC model to examine the microstructure change due to the laser cutting process, three observations were made. First, the largest grain growth occurs at the edge of the laser where the material was not hot enough to be cut. Then, grain growth decays as the distance from the edge increases. Finally, at the edge of the HAZ boundary, grain growth does not occur.

Browse

Recent Submissions