41530 research outputs found
Sort by
Synthesis and Testing of a Biodegradable Composite from Agricultural Waste
Get PDFIn this study, polylactic acid (PLA) and ground pistachio shells are combined into a biocomposite with 15 wt% shells through compression molding. To improve interfacial bonding, shells were treated by alkali cleaning and functionalized with tetraethoxysilane before processing. Two biocomposite sample types were fabricated: PLA/shells and PLA/shells/chemical additives. The chemical additives (maleic anhydride and dicumbyl peroxide) were investigated to improve matrix and reinforcement bonding. Difficulties in processing resulted in four testable samples. Mechanical property testing showed that plain PLA samples performed the best on average. Incorporation of the additives increased both ultimate tensile strength and Young’s modulus by 50% as compared to shells without additives. DSC and FTIR showed that thermal properties and chemical bonds were as expected. Overall, more samples are necessary to produce statistically accurate claims
Filling Gaps in Scientific Data Sets Using Physics Informed Neural Networks: A Case Study in Velocity Fields
Get PDFGaps in scientific data sets are a persistent issue for researchers in a variety of fields, and while nothing makes up for missing out on real data, well-simulated synthetic data can be a useful tool. In the world of image processing, machine learning techniques have become quite sophisticated at taking an image with a missing component and filling in that space with something believable. The aim of this thesis is to take machine learning techniques similar to what gets used in image processing and repurpose them to infill gaps in scientific data sets in a realistic manner. This thesis compares and contrasts some machine learning techniques, such as basic neural networks, convolutional neural networks, and autoencoders, before focusing on neural networks. Two approaches were considered for this problem: an approach inspired by the idea of taking the weighted average of a point\u27s neighbors to impute that point and an approach grounded in spatial coordinates. Then, a physics informed component is added to both of these networks to enforce certain physical realism standards. The physics informed component makes the neural networks marginally more accurate but noticeably more realistic and physically possible. Each of the networks, both before and after the physics informed component is added, is tested on a variety of velocity fields. Finally, a statistical analysis is done on the networks as compared to less sophisticated infill methods to demonstrate that they do perform well
Automation of Post Fermentation Must Removal
Get PDFThe Harvest Haulers project addresses a critical operational inefficiency at Saucelito Canyon Winery, where post-fermentation must removal from wine barrels was labor-intensive and potentially hazardous. This project aimed to develop a custom forklift attachment that could securely handle Bordeaux and Burgundy barrels, streamline the dumping process, and improve worker safety.
The resulting solution is a forklift-compatible fixture designed to lift, secure, and tilt barrels using a robust combination of a modified aluminum pallet, padded hoop, ratchet straps, and a custom hinge mechanism. The design meets all engineering requirements, including a 600 lb. load capacity, 135° tilt, and a single-operator setup under ten steps. Through extensive field testing and Finite Element Analysis (FEA), the fixture demonstrated excellent structural integrity and usability.
The system passed both empty and full-barrel operational tests. It achieved minimal deformation, maintained safe barrel alignment via precision-welded nubs, and required minimal maintenance. At a cost of just over 450 per unit.
Incorporating OSHA-compliant safety protocols, food-grade materials, and sustainability measures, this fixture meets both the functional and ethical standards required in the winemaking industry. The design has been validated by stakeholders, with recommendations for minor manufacturing improvements, making it ready for real-world application
TrueHue: Finding the Right Shade for You
Get PDFSix in ten women struggle to find the right foundation or concealer due to the lack of representation and unreliable color-matching tools. This highlights the need for a more inclusive solution in the beauty industry. Introducing TrueHue—an innovative app prototype designed to address the widespread challenge of accurate color matching. TrueHue leverages Convolutional Neural Network (CNN) AI technology to provide precise, personalized makeup shade recommendations. Users begin by selecting their skin type, desired price range, and type of product they’re seeking. The app then captures a full-face image to analyze undertones and overall skin tone. TrueHue recommends the best shade match and presents the top three matching products from a globally sourced database of makeup brands. Users can filter results by personal preferences and make purchases directly through the app—offering a seamless, inclusive, and empowering makeup experience for all
Subalpine and Alpine Plant Community Turnover in Yosemite National Park Following 30 Years of Climate Warming
Get PDFSubalpine and alpine regions in California are warming more rapidly than lower elevations. In response, high-elevation plants may acclimate or adapt to the new climate, move to higher elevations or latitudes, persist in refugia, or go extinct. While models predict substantial declines in subalpine and alpine plant communities due to discontinuous population distributions and lack of upslope area, actual responses may vary due to lags in dispersal, establishment, and extinction. Although shifts in high-elevation plant communities have been documented in other regions of the world, such changes have not yet been studied in Yosemite National Park. To assess high-elevation plant community shifts in Yosemite National Park, we resurveyed historic vegetation plots spanning 2,800 to 3,800 meters from conifer forests to alpine fell-fields. Over the past 30 years, in the treed zone, species richness declined by 6%. In the treeless zone, vegetation cover increased by 31% and species richness increased by 12%. Community composition shifted slightly towards lower-elevation assemblages, with no significant change in community climate affinity. Notably, Pinus albicaulis appears to be transitioning from krummholz to upright growth, with 41.1% of previously alpine plots now supporting upright trees. These trends may indicate early stages of ecosystem transition under climate change