Portion Cup Dispensing Module Forensic Report

I. Project Summary

• Role: Senior Mechanical Engineer (Project “Bender” sub-system).
• Mandate: Automate the singulation and placement of portion cups for high-volume digital order fulfillment — eliminating order defects and cross-contamination at sub-second dispensing precision.
• Core Achievement: ±2% portion accuracy across variable fluid viscosities against an industry standard of ±15%, sustained at 350 meals per hour with sensor-verified dispense integrity.

II. The Anatomy of Failure

1. Drive Train Crisis: Polymer Teeth Under Steel Torque

• Trigger: The 218-tooth internal gear was specified in Acetal (Delrin) for noise and self-lubrication — but the thermal load of 350 meals/hour against the aluminum enclosure created expansion-mismatch binding risk, and the torque required to drive viscous ingredients through the stainless helix shaft threatened to strip the polymer teeth under the steel pinion.
• Intervention: Enforced concentricity with cast Steralloy sleeve alignment guides between the drive train and the acetal helix, locked axial migration with low-clearance retaining rings, and accelerated auger-geometry validation by prototyping in Rigid 10K resin to mimic metal stiffness before hard tooling.
• Result: The reinforced module sustained the 350 meals/hour target at 99.9% order accuracy, decoupling portioning precision from fluid viscosity.

2. The Singulation Event: One Cup, Every Time

• Trigger: Nested cups resist clean separation — release two and the order is wrong; release zero and the system “ghost dispenses” sauce into a void it believes is a cup.
• Intervention: A helix-driven screw separation: rotating acetal helices engage the bottom cup’s rim, forcing it down while supporting the cup above — a mechanical handoff that releases exactly one. Stainless rings and a stabilizing rod hold axial alignment through the separation event, and every drop is verified by an optical drop sensor with a second photoelectric sensor watching stack height.
• Result: Closed-loop dispensing — the controller inhibits actuation on an empty stack instead of dry-firing, and inventory decrements only on a physically confirmed drop.

3. The Tilt Ramp: Holding Dinner Against Gravity

• Trigger: The positioner arms that tilt bowls to receive ingredients face backdrive torque: a spur-gear drive would need continuous motor current to hold a loaded bowl at angle — overheating coils in normal use and dumping the meal on any power loss.
• Intervention: Specified a self-locking worm-gear drive with a lead angle under 4°, where flank friction exceeds the gravity vector’s tangent: a mechanical lock requiring zero holding power. Absorbed the worm’s axial thrust with phosphor-bronze oil-embedded thrust washers, shielded ball bearings, machined retainers, and a 0.1–0.5 mm shim stack eliminating axial play.
• Result: Zero-energy static hold — the bowl stays put through E-stops and outages, with the efficiency trade-off (45–55%) accepted as the price of fail-safe load holding.

4. The Watchdog Layer: Detecting Silent Failures

• Trigger: In an unattended food cell, the dangerous failures are the quiet ones — a stalled gear cooking a motor, a missing bowl catching dispensed food, an ingredient bin drifting into the pathogen danger zone.
• Intervention: Layered sensing tied to the industrial controller: Hall-effect geartooth sensors with true-zero-speed capability flag stalls by correlating motor current against pulse trains; retro-reflective sensors confirm a vessel is present before any dispense; EtherCAT thermal inputs log hot (125–180°F) and cold zones to the millisecond for live HACCP compliance.
• Result: Jam conditions stop the drive before gear stripping, void dispenses are physically impossible, and food-safety state is continuously audited rather than spot-checked.

III. Governance & Rhythm

• The Pulse: High-velocity iteration on Onshape with a fleet of nine in-house SLA printers — auger and hopper geometry cycles compressed from two weeks to three days — backed by rapid-turn CNC and casting vendors.
• The Compliance Frame: NSF/ANSI 169 (special-purpose food equipment), UL 763 and UL 508A, NSF 51 food-zone materials (316 stainless shafts, acetal wear parts), IP69K washdown targets, and toolless disassembly in under 15 minutes for sanitation.
• The Artifacts: The expanded structured BOM (June 2021); the prototype buy list ($18,472 for four Rev 1 units); the published patent application documenting the chute, positioner, and sensor architecture.

IV. Quantified Impact

• Achieved ±2% portion accuracy across variable viscosities — versus the industry’s ±15%.
• Sustained 350 meals per hour at 99.9% order accuracy.
• Compressed prototype iteration from two weeks to three days via the in-house print fleet.
• Eliminated dry-fire and ghost-dispense failure modes through dual-sensor closed-loop verification.
• Engineered a zero-energy static hold on the tilt mechanism via self-locking worm geometry.
• Validated the module within an NSF-169 / UL 763 / UL 508A certification frame.

V. Source Trail

The claims above rest on the project’s primary evidence archive — structured BOMs, datasheets, and patent figures — compiled through the NotebookLM forensic registry:

• BOM_PORTION_CUP_DISPENSER_structured_expanded_06_21_2021 — component hierarchy, materials, and sensor line items.
• Portion Cup Prototype Buy List - Master BOM — the Rev 1 prototype cost basis.
• portion_cup.pdf / portion_cup_2.pdf — exploded views of the gear train and helix engagement.
• US20240164588A1 — the patent application covering chute, positioner arms, and sensing.
• SP17D020-20210710A — the tilt-ramp prototype quotation (worm housing, driven arm).
• HYPHEN-SPECSHEET.pdf — certifications and performance validation.
• 04028VA_IP69K.pdf — washdown-rated thermal management specs.