Aidan P. Steins

Summary

Results

Summary

Carbon fiber reinforced polymer (CFRP) patches are widely used across industries to repair cracked metallic and composite structures. Using past experimental research as a guide for our setup and expected results, the goal of this project was to develop a finite element model in ANSYS Mechanical of a pre-cracked metallic dogbone coupon reinforced with a CFRP patch under tensile loading (mimicking an Instron tensile test). We investigated the effect of three parameters—patch shape, number of layers, and adhesive thickness—on the maximum von Mises stress at the crack tip. We developed a full-factorial Design of Experiments (DOE) to minimize the number of trials required to quantify the effect of these three factors on the maximum stress at the crack tip. To avoid complex plastic material behavior, we performed all simulations within the linear elastic region of the involved materials. To minimize element count and computational time, 3D shell elements were used for all bodies in the model. The CFRP patches were modeled using ANSYS's 'Layered Section' feature (as shown in the first image collage) to reduce the number of contact definitions and layers required to model the rectangular, triangular, and inverted triangular patches. Boundary conditions were applied to mimic the grip and loading conditions of a real-world tensile test. Convergence studies were performed on each simulation by refining the mesh at the crack tip using a pre-defined methodology.

Results

The initial regression model included all main effects, two-way interactions, and three-way interactions; however, doing so left no degrees of freedom for error. After removing the three-way interactions from the model, we were left with the main effects and two-way interaction terms. After reviewing the normal probability and fitted-value plots for this model, we determined that a linear model adequately captured the effects of the three factors on the maximum stress at the crack tip, as confirmed by a high R-squared value. Shown in the second image collage are the main effects and interaction plots using this reduced linear model. We can see that the number of composite layers has the largest effect on maximum stress at the crack tip, followed by patch shape, and lastly adhesive thickness. It is important to note that adhesive thickness may have a more significant impact on maximum stress at the crack tip if more elements were available to model through the thickness of the adhesive layer. The interaction plots show that the interaction between the number of composite layers and patch shape, as well as the interaction between layer count and adhesive thickness, are statistically significant. However, because adhesive thickness alone is statistically insignificant in this model, the interaction between layer count and adhesive thickness is less critical than the interaction between layer count and patch shape. From these conclusions, we created our third and final reduced model, which retained only the statistically significant factors and interactions.

Summary

As part of an industry project, our client needed a Digital Imaging Technician (DIT) workstation and portable power station capable of supporting a film set’s power needs and EV charging in both local and remote locations. Our team retrofitted in-house designed and manufactured components to an all-electric SUV that fulfilled the needs of our client’s technicians. I contributed to the design and documentation, and supported the fabrication of the DIT’s internal workstation, external viewing hatch, and retractable server rack, ensuring proper ergonomics and overall function. The power station trailer frame houses four battery modules and supports a fully deployable and articulating solar array which can be precisely adjusted to harness maximum solar power. Half of the array's panels are always exposed, allowing for power generation during transportation or in confined spaces. The trailer can provide EV charging and support almost any power-demanding equipment. I was primarily responsible for the conceptual structure design, component sourcing, and hand calculations required for proper actuation of the array. I also assisted in the CAD modeling, documentation, and fabrication of the array’s structure and components.

Results

We received positive ergonomic and usability feedback on the DIT vehicle from a variety of users, specifically for the internal workstation and the external viewing hatch. All subsystems performed as expected and remained functional after transportation. Preliminary tests of the solar array proved that it articulated reliably with continued use, which validated the analytical hand calculations. The precise angle adjustability of the array also allowed for maximum solar power generation. Overall, the DIT vehicle and solar array concept proved successful and both have the opportunity for future feature enhancements.

Summary

An initial prototype of a portable, isobutane camp-stove-powered water heater is shown in the project gallery. It utilizes primary and secondary heat exchangers that mimic the setup of a convection-based tankless water heater. Cold water first flows through the secondary heat exchanger, positioned above the primary, where it condenses water from the stove's exhaust gases onto copper coils, releasing latent heat to the cool water within. The water then passes through the primary heat exchanger, which provides most of the convective heat transfer. A CFD model was developed to simulate the heat transfer from the stove's flame to the copper tube and water. After validation against physical experimental testing, the model helped visualize heat transfer and optimize the coil size and orientation. It will also assist in improving overall efficiency when implementing design changes in the future.

Results

On average, the steady-state temperature difference (ΔT) between the water at the inlet and outlet was roughly 32° F at a flow rate of 1.13 L/min. Based on the expended energy of the camp stove, determined by the amount of fuel used over a given period of time, the efficiency of the device was calculated to be 70.5%.

Summary

The goal of this project was to develop a speech recognition system for veterans with missing digits or numbness in their hands, enabling them to shift gears on their bikes. The device needed to recognize the wake word, “marvin”, as well as the keywords “up” and “down”, and communicate this information wirelessly to a wireless derailleur, which would shift the bike accordingly. The device utilized a neural decision processor trained on thousands of voice samples of the aforementioned keywords. A water-resistant, resin-printed enclosure housing the battery, microcontroller, and gooseneck-style microphone was attached to the rider's helmet. Upon identifying an up or down shift command, the device transmits a packet via BLE to the shifter, which executes a shift in the bike’s gears.

Results

After testing the device multiple times with four different users, the device's accuracy was found to be approximately 75%, with a peak of 85%. However, as each user used the device more, the accuracy improved, likely because they adjusted the way they spoke their commands to better match what the device was trained to recognize.