Five Roles.
One Throughline.

I currently conduct undergraduate research on wire-laser Directed Energy Deposition (DED), supported by a $4,500 McCormick Summer Undergraduate Research Award (the Robert and Barbara Feldmann Undergraduate Research Award) - funding I secured by developing an independent research proposal and experimental plan rather than simply joining an existing lab workflow. The work spans two research streams and sits at the intersection of mechanical design, fluid dynamics, materials science, vibrations, and scientific software development.

This stream investigates whether controlled low-frequency mechanical vibration applied to the substrate during wire-laser DED of Inconel 718 can change melt-pool fluid motion, solidification behavior, and grain morphology - specifically, whether vibrating near a relevant natural frequency can promote a more homogeneous, equiaxed grain structure instead of the directional, columnar structure conventional DED tends to produce. The experimental matrix spans a baseline case plus 50, 200, 350, and 500 Hz.

A large share of this work has been mechanical: designing a vibration fixture that can transmit lateral motion to a small metal substrate inside a laser deposition machine without disturbing the rest of the system. Early concepts that clamped directly to the machine's existing rigid jaw failed for a simple reason - a rigidly constrained plate absorbs the exact displacement the exciter is trying to create. The concept that solved this uses a fixed aluminum base plate, a lighter moving plate connected through thin spring-steel flexures, and an accelerometer positioned to measure the substrate's actual motion rather than assuming the commanded signal reached it unchanged.

In parallel, I'm building a multiphysics melt-pool model in ANSYS Fluent using Volume of Fluid multiphase modeling and custom C user-defined functions to implement a moving Gaussian laser heat source, wire mass/momentum addition, surface-tension-driven (Marangoni) flow, and vibration forcing. A major share of this work has been diagnosing numerical instability - Courant-number spikes, reverse flow at pressure outlets, and restart inconsistencies - and learning to distinguish a genuine physical result from a numerical artifact, ultimately converging on a stable adaptive time-step range between 5×10⁻⁷ s and 5×10⁻⁵ s.

I'm also looking into cavity and bubble formation in wire-laser DED, analyzing high-speed imaging to understand how cavities initiate, grow, and collapse, and how that behavior relates to bubble generation and potential trapped porosity - a mechanism that remains poorly characterized in this process. My contribution includes frame-by-frame image analysis and developing a custom analysis tool to make that process faster and more consistent: bubble tracking with nearest-neighbor linking, polygon-based cavity boundary tracing, image contrast/ROI enhancement, and a unified data export format so a full analysis session can be resumed, reviewed, or shared.

At Tackform, I moved into product-focused mechanical design - working on vehicle-mounted accessories and mounting hardware, including device mounts, fairleads, and mounting plates, designed and revised in SolidWorks. Around 15 manufacturable components moved through my hands over the internship, each intended for real automotive use, not just a rendering.

Because these parts live on a moving vehicle, I had to design around vibration, repeated loading, and fastener load paths - not just static loading. I also applied DFMA principles throughout, since every part needed to actually be manufactured and assembled reliably by a third-party supplier, not just look correct in CAD.

My first plant-floor engineering role, supporting the K12 cold-end line at a fiberglass insulation facility generating more than $300,000 per day in product value - an environment where a small equipment failure could create significant downtime fast. I worked directly with nodulators, automated baggers, compression rams, conveyors, robotic stacking systems, and date-coding equipment, performing root-cause analysis on recurring mechanical and pneumatic failures.

One of the most important lessons of the internship had nothing to do with equipment drawings: operators and maintenance technicians who worked the line every day had years of pattern recognition that no manual could replace. I learned to treat their input as real engineering evidence, not just anecdotal color layered on top of "real" analysis.

My primary project addressed a recurring date-coding and product-traceability issue on the line. I studied the physical layout, product path, and operating sequence, gathered input from operators and technicians who dealt with the problem daily, and developed a revised date-coder placement in OnShape - incorporating a UV detector and photocell arrangement, and accounting for maintenance access and the plant's management-of-change process. I coordinated the redesign's implementation across four bagger lines.

The result: approximately a 5% reduction in downtime, recovering roughly $2,000 per day in production output.

I also supported approximately $2.2 million in bagger-system capital upgrades - reviewing equipment layouts, sourcing components in SAP, comparing vendor quotations, and evaluating installation, operator access, and maintenance serviceability to make sure new equipment would actually integrate with existing plant operations. I gained hands-on experience with lockout/tagout, management of change, and 5S, and learned firsthand that safety, guarding, and serviceability have to be designed in from the start - not added on after the fact.

I returned to City National Bank for a second summer in a more technical role, supporting workforce planning, cost modeling, client-data analysis, and business intelligence reporting. I built and maintained Excel-based workforce and cost models used to evaluate staffing scenarios, and used SQL to extract, merge, clean, and validate client data - then used SQL and Qlik to support client segmentation and reporting across the bank's data systems.

The biggest lesson of the summer: a query running successfully and a query being correct are two different things. I cleaned, segmented, and validated hundreds of client records, checking for missing information, inconsistent classifications, duplicates, and logic that produced results that didn't actually make business sense. That validation mindset - never assuming an output is right just because the tool ran without errors - carried directly into how I approach simulation and experimental data today.

My first professional internship was outside mechanical engineering entirely, but it built a foundation I still use: turning raw performance data into a decision someone can actually act on. I evaluated the performance of individual bank branches, diagnosed why certain locations were underperforming, and helped build structured turnaround plans - considering operating costs, revenue, employee productivity, and branch efficiency against internal targets.

The real skill wasn't spotting that a number was low - it was distinguishing symptoms from root causes, prioritizing which gap actually mattered most, and packaging findings into a sequence of actions a branch manager could implement, using Excel as the central tool for organizing and presenting the analysis.