The subfab was a 3-foot vertical envelope below the cleanroom. Every utility run (UPW, caustic chemicals, nitrogen) had to route through this space. The hardest part wasn't the design itself but constant awareness of access: making sure there was clearance for tooling, for the installers and mechanical crews who'd actually build and maintain these systems. In semiconductor, leak prevention, detection, and containment aren't optional. A chemical leak in a confined subfloor is catastrophic. Every design decision started with "can someone get in here to fix this?"

Standardization meant any crew could install any rack without custom instructions, which cut down on errors and installation time across the fab. When you're running dozens of identical utility drops, consistency isn't just convenient, it's how you prevent the kind of mistakes that shut down a cleanroom.

A mix of ground-up and optimization work. For the greenfield project, I developed costing specs, translating process intent into mechanical scope with real numbers before anything was built. The kind of early-stage work where getting the specs wrong means the budget is fiction.

Conducted a plant-wide heat exchanger efficiency study, comparing design vs. actual performance to flag energy recovery opportunities. The wins are in the details: fouled exchangers running at 60% efficiency, oversized pumps burning electricity, bypass valves stuck open for years. A 5% efficiency gain on a single exchanger can save six figures annually.

Tight timeline, tight coordination with manufacturing and purchasing. The kind of project where one missed tolerance or late vendor delivery cascades through the whole schedule. Managed every phase from initial design through final shipment.

The company had just invested in a brand new large-format CNC machine. The challenge was machining very large steel weldments, extremely heavy parts that needed accuracy across the full machine envelope. I designed a custom fixture that allowed shifting weldments across the full table without lifting and resetting. Without it, every repositioning meant craning out the weldment, re-aligning, re-zeroing. Hours of downtime per piece.

The core challenge was vibration. Running dynamic calculations for K values and damper selection to ensure high-speed production equipment stayed stable under load. The machines operate at production speeds where resonance is a real risk. Get the damping wrong and a $2M machine becomes a very expensive blender. Proper damper selection is the difference between a machine that runs for years and one that vibrates itself to pieces in months.

Managed paving projects across Oregon, working directly with foremen and ODOT. The biggest was overnight paving along the I-5 corridor north of Salem. Shutting down lanes at 10 PM, everything paved, striped, and open by morning rush. I managed costing reports, coordinated vendors, met with stakeholders, and when things got tight, drove equipment to site myself. Also handled A/P, A/R, and project takeoffs. The kind of work where you learn fast that a plan is only as good as the people executing it.