Edge of PracticeShort-Cycle FalsificationProcess Control Boundary

Gauge-Correlated Asymmetry in Polymer Cooling

Cooling symmetry is admissible only if downstream product artifacts remain invariant under controlled variation of cooling time. Reproducible, gauge-correlated defects invalidate the symmetry assumption for operational use.

Core Doctrine

A process symmetry assumption is admissible only if controlled parameter variation does not produce reproducible, product-correlated asymmetry. Otherwise, the assumption is operationally void.

Boundary Summary
Valid only if

No reproducible artifact trend appears across cooling-time variation.

Invalid when

A monotonic or directional defect signature correlates with cooling time.

Governing variable

Product-correlated artifact response to controlled cooling variation.

Tested Assumption

Cooling suppresses operational asymmetry

Standard cooling times are assumed to suppress internal thermal asymmetry such that no reproducible downstream defect emerges under nominal conditions.

Why This Assumption Persists

Throughput and invisibility

  • Fixed cooling times maximize throughput
  • Internal gradients are not directly observable
  • Downstream defects are averaged within tolerance
  • Low scrap rates are misinterpreted as validation
Falsification Protocol

Minimal plant-ready experiment

  • Standard PP part (~4 mm wall)
  • All parameters fixed except cooling time
  • Three conditions: nominal, −20%, +20%
  • 30 parts per condition

No sensors. No tooling changes. No instrumentation dependency.

Single Primary Readout

One artifact, consistently measured

  • Warpage magnitude/direction
  • Sink depth
  • Shrink differential
  • Optical banding

Multiple metrics are not admissible. One signal must carry the test.

PASS

No systematic trend across cooling-time variation.

FAIL

Reproducible defect shifts with cooling time.

What can be shifted by control is not suppressed by assumption.

If a process parameter produces reproducible product asymmetry, the symmetry assumption is no longer admissible for control decisions.

Part of the Edge of Practice short-cycle experiment index