Matching Items (214)
Description
The paper presents a new exhaust header design to replace the current design on Arizona State University's Formula SAE car. Also, the thought process of the design was presented as well as a method of analysis for tuning the exhaust headers. The equation presented was then compared with a computational fluid dynamics model using ANSYS Fluent. It was found that the equation did not match the timing of the CFD model. However, the design does allow for simple changes to be made in order to reduce the length of the exhaust and allow for the correct tuning. Also, the design minimizes interference between the individual headers which is ideal to increase engine performance. The exhaust meets the Formula SAE regulations, and is designed to fit in the new chassis for the FSAE car that ASU will run in 2015. Recommendations were also made to further improve the design and analysis model.
ContributorsKaashoek, Kevin Jason (Author) / Huang, Huei-Ping (Thesis director) / Trimble, Steven (Committee member) / Barrett, The Honors College (Contributor) / Mechanical and Aerospace Engineering Program (Contributor)
Created2014-05
Description
This dissertation investigates the potential for enzyme induced carbonate cementation as an alternative to Portland cement for creating building material from sand aggregate. We create a solution of urease enzyme, calcium chloride (CaCl2), and urea in water and added sand. The urease catalyzes the synthesis of carbonate from urea, and the carbonate then bonds with a dissociated calcium ion and precipitates from the solution as calcium carbonate (CaCO3). This precipitate can form small crystal bridges at contacts between sand grains that lock the sand grains in place. Using enzyme induced carbonate precipitation we created a cemented sand sample with a maximum compressive strength of 319 kPa and an elastic modulus of approximately 10 MPa. Images from the SEM showed that a major failure mechanism in the cemented samples was the delamination of the CaCO3 from the sand grains. We observed that CaCO3 cementation did not when solutions with high concentrations of CaCl2 and urea were used.
ContributorsBull, Michael Ryan (Author) / Kavazanjian, Edward (Thesis director) / Chawla, Nikhilesh (Committee member) / Barrett, The Honors College (Contributor) / Materials Science and Engineering Program (Contributor)
Created2014-05
Description
This study examines the impact of spatial landscape configuration (e.g., clustered, dispersed) on land-surface temperatures (LST) over Phoenix, Arizona, and Las Vegas, Nevada, USA. We classified detailed land-cover types via object-based image analysis (OBIA) using Geoeye-1 at 3-m resolution (Las Vegas) and QuickBird at 2.4-m resolution (Phoenix). Spatial autocorrelation (local Moran’s I ) was then used to test for spatial dependence and to determine how clustered or dispersed points were arranged. Next, we used Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data acquired over Phoenix (daytime on 10 June and nighttime on 17 October 2011) and Las Vegas (daytime on 6 July and nighttime on 27 August 2005) to examine day- and nighttime LST with regard to the spatial arrangement of anthropogenic and vegetation features. Local Moran’s I values of each land-cover type were spatially correlated to surface temperature. The spatial configuration of grass and trees shows strong negative correlations with LST, implying that clustered vegetation lowers surface temperatures more effectively. In contrast, clustered spatial arrangements of anthropogenic land-cover types, especially impervious surfaces and open soil, elevate LST. These findings suggest that city planners and managers should, where possible, incorporate clustered grass and trees to disperse unmanaged soil and paved surfaces, and fill open unmanaged soil with vegetation. Our findings are in line with national efforts to augment and strengthen green infrastructure, complete streets, parking management, and transit-oriented development practices, and reduce sprawling, unwalkable housing development.
ContributorsMyint, Soe Win (Author) / Zheng, Baojuan (Author) / Talen, Emily (Author) / Fan, Chao (Author) / Kaplan, Shari (Author) / Middel, Ariane (Author) / Smith, Martin (Author) / Huang, Huei-Ping (Author) / Brazel, Anthony J. (Author)
Created2015-06-29
Description
Passive flow control achieved by surface dimpling can be an effective strategy for reducing drag around bluff bodies - an example of substantial popular interest being the flow around a golf ball. While the general effect of dimples causing a delay of boundary layer separation is well known, the mechanisms contributing to this phenomena are subtle and not thoroughly understood. Numerical models offer a powerful approach for studying drag reduction, however simulation strategies are challenged by complex geometries, and in applications the introduction of ad hoc turbulence models which introduce additional uncertainty. These and other factors provide much of the motivation for the current study, which focused on the numerical simulations of the flow over a simplified configuration consisting of a dimpled flat plate. The principal goals of the work are to understand the performance of the numerical methodology, and gain insight into the underlying physics of the flow. Direct numerical simulation of the incompressible Navier-Stokes equations using a fractional step method was employed, with the dimpled flat plate represented using an immersed boundary method. The dimple geometry utilizes a fixed dimple aspect ratio, with dimples arranged in a single spanwise row. The grid sizes considered ranged from approximately 3 to 99 million grid points. Reynolds numbers of 3000 and 4000 based on the inlet laminar boundary layer thickness were simulated. A turbulent boundary layer was induced downstream of the dimples for Reynolds numbers which did not transition for the flow over an undimpled flat plate. First and second order statistics of the boundary layer that develops agree reasonably well with those for turbulent channel flow and flat plate boundary layers in the sublayer and buffer layers, but differ in the outer layer. Inspection of flow visualizations suggest that early transition is promoted by thinning of the boundary layer, initiation of shear layer instabilities over the dimples, flow separation and reattachment, and tripping of the boundary layer at the trailing edge of the dimples.
ContributorsMode, Jeffrey Michael (Author) / Squires, Kyle (Thesis advisor) / Herrmann, Marcus (Committee member) / Huang, Huei-Ping (Committee member) / Arizona State University (Publisher)
Created2010
Description
This study explores the relationship between three physics-based predictive models defined by Castruccio et al. (2013), and four different distinct experimental morphologies of lava flows produced in a series of laboratory simulations where polyethylene glycol 600 (PEG) was pumped into an inclined chilled bath of water. The length of the experimental flow was recorded over time to create an experimental model to later be compared to the physics-based predictive models. The experimental morphologies are pillowed, rifted, folded, and leveed flows which can be characterized by a dimensionless parameter 𝛹, which scales natural lava flows to experimental lava flows and is a ratio of timescales, the characteristic timescale of thermal flux from the vent and the characteristic timescale of crust formation caused by surface cooling (Fink and Griffiths 1990). The three physics-based models are presented such that the downslope gravitational acceleration drives the flow, while either the Newtonian viscosity of the flow, the Yield Strength of the core (YS), or the Yield Strength of the growing crust (YSC) is the primary retarding factor in flow propagation. This study concluded that low 𝛹-value flows (low flux, low temperature, extensive crust formation) are better captured by the YSC model. And although the Newtonian model did not perfectly capture the behavior of any experimental flows in this study, high 𝛹-value flows (high flux, high temperature, little crust formation) that formed levees exhibited the most Newtonian behavior.
ContributorsCourtney, Cara Alexandra (Author) / Clarke, Amanda B. (Thesis director) / Huang, Huei-Ping (Committee member) / Williams, David A. (Committee member) / School of Sustainability (Contributor) / Mechanical and Aerospace Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2020-05
Description
An understanding of aerodynamics is crucial for automobile performance and efficiency. There are many types of “add-on” aerodynamic devices for cars including wings, splitters, and vortex generators. While these have been studied extensively, rear spoilers have not, and their effects are not as widely known. A Computational Fluid Dynamics (CFD) and wind tunnel study was performed to study the effects of spoilers on vehicle aerodynamics and performance. Vehicle aerodynamics is geometry dependent, meaning what applies to one car may or may not apply on another. So, the Scion FRS was chosen as the test vehicle because it is has the “classic” sports car configuration with a long hood, short rear, and 2+2 passenger cabin while also being widely sold with a plethora of aftermarket aerodynamic modifications available. Due to computing and licensing restrictions, only a 2D CFD simulation was performed in ANSYS Fluent 19.1. A surface model of the centerline of the car was created in SolidWorks and imported into ANSYS, where the domain was created. A mesh convergence study was run to determine the optimum mesh size, and Realizable k-epsilon was the chosen physics model. The wind tunnel lacked equipment to record quantifiable data, so the wind tunnel was utilized for flow visualization on a 1/24 scale car model to compare with the CFD.
0° spoilers reduced the wake area behind the car, decreasing pressure drag but also decreasing underbody flow, causing a reduction in drag and downforce. Angled spoilers increased the wake area behind the car, increasing pressure drag but also increasing underbody flow, causing an increase in drag and downforce. Longer spoilers increased these effects compared to shorter spoilers, and short spoilers at different angles did not create significantly different effects. 0° spoilers would be best suited for cases that prioritize fuel economy or straight-line acceleration and speed due to the drag reduction, while angled spoilers would be best suited for cars requiring downforce. The angle and length of spoiler would depend on the downforce needed, which is dependent on the track.
0° spoilers reduced the wake area behind the car, decreasing pressure drag but also decreasing underbody flow, causing a reduction in drag and downforce. Angled spoilers increased the wake area behind the car, increasing pressure drag but also increasing underbody flow, causing an increase in drag and downforce. Longer spoilers increased these effects compared to shorter spoilers, and short spoilers at different angles did not create significantly different effects. 0° spoilers would be best suited for cases that prioritize fuel economy or straight-line acceleration and speed due to the drag reduction, while angled spoilers would be best suited for cars requiring downforce. The angle and length of spoiler would depend on the downforce needed, which is dependent on the track.
ContributorsNie, Alexander (Author) / Wells, Valana (Thesis director) / Huang, Huei-Ping (Committee member) / Mechanical and Aerospace Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2019-12
Description
Aluminum alloys are commonly used for engineering applications due to their high strength to weight ratio, low weight, and low cost. Pitting corrosion, accelerated by saltwater environments, leads to fatigue cracks and stress corrosion cracking during service. Two-dimensional (2D) characterization methods are typically used to identify and characterize corrosion; however, these methods are destructive and do not enable an efficient means of quantifying mechanisms of pit initiation and growth. In this study, lab-scale x-ray microtomography was used to non-destructively observe, quantify, and understand pit growth in three dimensions over a 20-day corrosion period in the AA7075-T651 alloy. The XRT process, capable of imaging sample volumes with a resolution near one micrometer, was found to be an ideal tool for large-volume pit examination. Pit depths were quantified over time using renderings of sample volumes, leading to an understanding of how inclusion particles, oxide breakdown, and corrosion mechanisms impact the growth and morphology of pits. This process, when carried out on samples produced with two different rolling directions and rolling extents, yielded novel insights into the long-term macroscopic corrosion behaviors impacted by alloy production and design. Key among these were the determinations that the alloy’s rolling direction produces a significant difference in the average growth rate of pits and that the corrosion product layer loses its passivating effect as a result of cyclic immersion. In addition, a new mechanism of pitting corrosion is proposed which is focused on the pseudo-random spatial distribution of iron-rich inclusion particles in the alloy matrix, which produces a random distribution of pit depths based on the occurrence of co-operative corrosion near inclusion clusters.
ContributorsSinclair, Daniel Ritchie (Author) / Chawla, Nikhilesh (Thesis director) / Jiao, Yang (Committee member) / Bale, Hrishikesh (Committee member) / School of International Letters and Cultures (Contributor) / Materials Science and Engineering Program (Contributor, Contributor) / Barrett, The Honors College (Contributor)
Created2020-05
Description
In this paper, the effectiveness and practical applications of cooling a computer's CPU using mineral oil is investigated. A computer processor or CPU may be immersed along with other electronics in mineral oil and still be operational. The mineral oil acts as a dielectric and prevents shorts in the electronics while also being thermally conductive and cooling the CPU. A simple comparison of a flat plate immersed in air versus mineral oil is considered using analytical natural convection correlations. The result of this comparison indicates that the plate cooled by natural convection in air would operate at 98.41[°C] while the plate cooled by mineral oil would operate at 32.20 [°C]. Next, CFD in ANSYS Fluent was used to conduct simulation with forced convection representing a CPU fan driving fluid flow to cool the CPU. A comparison is made between cooling done with air and mineral oil. The results of the CFD simulation results indicate that using mineral oil as a substitute to air as the cooling fluid reduced the CPU operating temperature by sixty degrees Celsius. The use of mineral oil as a cooling fluid for a consumer computer has valid thermal benefits, but the practical challenges of the method will likely prevent widespread adoption.
ContributorsTichacek, Louis Joseph (Author) / Huang, Huei-Ping (Thesis director) / Herrmann, Marcus (Committee member) / Middleton, James (Committee member) / Mechanical and Aerospace Engineering Program (Contributor, Contributor) / Barrett, The Honors College (Contributor)
Created2016-12
Description
Due to high DRAM access latency and energy, several convolutional neural network(CNN) accelerators face performance and energy efficiency challenges, which are critical for embedded implementations. As these applications exploit larger datasets, memory accesses of these emerging applications are increasing. As a result, it is difficult to predict the combined dynamic random access memory (DRAM) workload behavior, which can sabotage memory optimizations in software. To understand the impact of external memory access on CNN accelerators which reduces the high DRAMaccess latency and energy, simulators such as RAMULATOR and VAMPIRE have been proposed in prior work. In this work, we utilize these simulators to benchmark external memory access in CNN accelerators. Experiments are performed generating trace files based on the number of parameters and data precision and also using trace file generated for CNN Accelerator Altera Arria 10 GX 1150 FPGA data to complete the end to end workflow using the mentioned simulators. Besides that, certain modifications were made in the default VAMPIRE code to implement certain functionalities such as PREA(Precharge All) and REF(Refresh). Then, precalculated energies were computed for DDR3, DDR4, and HBM based on the micron model to mention it in the dram specification file inputted to the VAMPIRE tool. An experimental study was performed and a comparison is made between DDR3, DDR4, and HBM, it was proved that DDR4 is nearly 31% more energy-efficient than DDR3 and HBMis 54% energy-efficient than DDR3. Performed modeling and experimental analysis on a large set of data and then split it into a set of data and compared the results of the small sets multiplied with the number of sets and the large data set and concluded that the results were nearly the same. Finally, a GUI is developed by wrapping both the simulators. GUI provides user-friendly access and can analyze the parameters without much prior knowledge and understanding of the working.
ContributorsPannala, Manvitha (Author) / Cao, Yu (Thesis advisor) / Chakrabarti, Chaitali (Committee member) / Seo, Jae-Sun (Committee member) / Arizona State University (Publisher)
Created2021
Description
Convolutional neural networks(CNNs) achieve high accuracy on large datasets but requires significant computation and storage requirement for training/testing. While many applications demand low latency and energy-efficient processing of the images, deploying these complex algorithms on the hardware is a challenging task. This dissertation first presents a compiler-based CNN training accelerator using DDR3 and HBM2 memory. An optimized RTL library is implemented to perform training-specific tasks and an RTL compiler is developed to generate FPGA-synthesizable RTL based on user-defined constraints. High Bandwidth Memory(HBM) provides efficient off-chip communication and improves the training performance. The impact of HBM2 on CNN training workloads is analyzed and compressively compared with DDR3. For training ResNet-20/VGG-like CNNs for the CIFAR-10 dataset, the proposed CNN training accelerator on Stratix-10 GX FPGA(DDR3) demonstrates 479 GOPS performance, and on Stratix-10 MX FPGA(HBM) shows 4.5/9.7 X energy-efficiency improvement compared to Tesla V100 GPU. Next, the FPGA online learning accelerator is presented. Adopting model segmentation techniques from Progressive Segmented Training(PST), the online learning accelerator achieved a 4.2X reduction in training latency. Furthermore, this dissertation presents an 8-bit floating-point (FP8) training processor which implements (1) Highly parallel tensor cores that maintain high PE utilization, (2) Hardware-efficient channel gating for dynamic output activation sparsity (3) Dynamic weight sparsity based on group Lasso (4) Gradient skipping based on FP prediction error. The 28nm prototype chip demonstrates significant improvements in FLOPs reduction (7.3×), energy efficiency (16.4 TFLOPS/W), and overall training latency speedup (4.7×) for both supervised training and self-supervised training tasks. In addition to the training accelerators, this dissertation also presents a CNN inference accelerator on ASIC(FixyNN) and FPGA(FixyFPGA). FixyNN consists of a fixed-weight feature extractor that generates ubiquitous CNN features and a conventional programmable CNN accelerator. In the fixed-weight feature extractor, the network weights are hard-coded into hardware and used as a fixed operand for the multiplication. Experimental results demonstrate FixyNN can achieve very high energy efficiencies up to 26.6 TOPS/W, and FixyFPGA achieves $2.34\times$ higher GOPS on ImageNet classification. In summary, this dissertation comprehensively discusses novel architectures of high-performance and energy-efficient ASIC/FPGA CNN inference/training accelerators.
ContributorsKolala Venkataramaniah, Shreyas (Author) / Seo, Jae-Sun (Thesis advisor) / Cao, Yu (Committee member) / Chakrabarti, Chaitali (Committee member) / Fan, Deliang (Committee member) / Arizona State University (Publisher)
Created2022