Industrial and commercial sectors continue to demand more power, requiring smaller and more efficient power converters. Wide-bandgap semiconductors like Gallium Nitride (GaN) Field Effect Transistors are capable of meeting higher rating benchmarks. However, an essential part of these converters, high-frequency magnetic materials, remains underdeveloped. This challenge is attributed to a fundamental lack of reliable data above 1 Mhz needed to characterize these materials adequately. A parameter used to characterize magnetic materials effectively, quality factor (Q), is the ratio of energy stored to energy lost. Although Q’s importance is recognized, there is an insufficient amount of data to develop magnetics. Existing measurement devices like Q-meter and LCR meters are either outdated, difficult to access, expensive, manually operated, or limited in their measuring capabilities. To close this gap, we propose a low-cost, automated, accessible, and accurate measurement device to drive high-frequency magnetic material research. The proposed Q-meter automatically measures across 1Mhz - 20Mhz to adequately characterize the magnetic material within 20 percent tolerance. An overall cost-effective device below 500 dollars will be widely accessible. Our approach allows researchers at any level to contribute effectively, without finding paywalls or requiring training.

Ultrafiltration membranes offer an energy-efficient, compact, and modular method to purify water, selectively separate pharmaceuticals and perform numerous separation processes. However, ultrafiltration membranes are often challenged with maintaining high permeability and selectivity, balancing good mechanical strength and transport properties, and resisting fouling, a phenomenon where debris accumulates onto membrane surfaces over time. Previously, block copolymer self-assembly has been used to achieve both high permeability and selectivity. Zwitterions, which are molecules containing positive and negative charges but have a net neutral charge, have been used to create a hydration layer on ultrafiltration membrane surfaces and introduce antifouling properties. This project aims to combine these strategies, namely zwitterions and block copolymer self-assembly, to create membranes with good mechanical strength, balanced transport properties, and fouling resistance. A zwitterion-containing block copolymer was synthesized using a two-step reversible addition fragmentation transfer polymerization that consisted of creating a zwitterionic, hydrophilic block of methyl methacrylate and 2-methacryloyloxyethyl phosphorylcholine and then adding on a mechanically robust styrene and acrylonitrile block via dispersion. Membranes were created from this polymer using non-solvent induced phase separation, a popular method for industrial membrane fabrication. Formulated batches of these zwitterionic block copolymers were analyzed, and the membranes’ performance has proven promising in their ability to create favorable water purification membranes, with initial permeances of 584 LMH/bar. This presentation aims to present the synthesis and characterization of these zwitterionic block copolymers, their subsequent fabrication into ultrafiltration membranes, and preliminary transport and mechanical data. Future research directions and immediate next steps will conclude the talk.

Rotating detonation rocket engines (RDREs) sustain an ultra-fast (10–100 kHz) detonation wave that circles an annular combustor, offering attractive efficiency gains over conventional burners. However, diagnostics become especially problematic with hydrogen–oxygen test fires: weak chemiluminescence and poor wall reflections yield low-contrast, high-speed videos where the wavefront melts into the background. Classic edge-detection techniques often fail, so frequency and wavenumber can’t be trusted. I’m addressing this gap by developing a lightweight convolutional neural network that pinpoints the combustor boundaries even in the dimmest footage so each frame can be unwrapped onto a one-dimensional annulus. I hand-curated 500 high-speed RDRE frames, annotating inner and outer combustor walls and training a pruned YOLOv4-Tiny detector in MATLAB; depth-wise-separable layers and custom anchors keep inference below 5 ms while preserving accuracy (AP@0.50 IoU = 0.88 for the outer wall, 0.52 for the inner, overall mAP = 0.70). The trained detector is inserted inline: each live frame is boxed, cropped, unwrapped to polar coordinates, and passed to existing brightness-trace/SVD routines that previously failed whenever low-contrast hydrogen-rich plumes obscured edges. Once masked, standard filtering and Fourier analysis recover wave speed and mode within minutes, turning raw pixels into reliable diagnostics between hot-fire runs. End-to-end testing shows the downstream wave-tracking algorithm now converges on the correct dominant mode in 97 % of frames versus 68 % with legacy edge detection. This approach also broadens the range of fuel-oxidizer mixtures that can be evaluated without resorting to computationally intensive CFD or other advanced simulations.

Greenhouse gases are produced from automobiles, aircraft, mines, AC, fires, animals, and rotting food. These gases disrupt the global climate, destroying wildlife, the environment, and crops. The transportation sector alone contributes 29% of emissions in the U.S, with one third of that from aircraft alone. If commercial planes were electric, we could reduce the amount of transportation emissions from the U.S, a study shows an estimated 93% reduction in emissions when replacing a jet from a Cessna 460XL with an electric one. This effectively gives insight into what every plane could be like. The primary challenge with developing electric commercial airplanes is overcoming the battery to weight ratio required for functionality. Determined to create a way to reduce the weight to energy ratio for a commercial plane battery, I’ve designed 3 different regenerative generators for the Airbus A320 wing. The generators use induction motor, Newton’s bladeless motor, and pneumatic motor. These generators convert wind that hits the plane midflight into electricity, recharging the battery, effectively reducing the need for a large battery. To test how effective these generators would be on a plane, I tested the weight difference, drag, and electrical efficiency of the wings with electric generators compared to a scaled down version of an unchanged A320 wing. These designs demonstrate some promising results for some of the designs and the possible future for electric planes.