Additive Manufacturing Quality Control Under Production Variability: Quantifying Dimensional Accuracy, Surface Integrity, and Process Stability Using In-Process and Post-Process Metrology

Abstract

This article proposes an applied engineering quality control framework that treats part quality as the output of a controlled process with measurable uncertainty rather than as a post hoc inspection event. The framework integrates in-process monitoring outputs with post-process metrology to quantify how variability propagates into dimensional deviation and surface roughness, and it formalizes decision governance using statistically defined acceptance criteria, false reject constraints, and escape risk limits. A production relevant tiered inspection strategy is evaluated conceptually through three inspection tiers: rapid in-process screening, intermediate dimensional sampling, and final verification metrology. Representative results patterns show that dimensional deviation is strongly location dependent within the build volume, surface roughness is disproportionately sensitive to build orientation and energy density, and anomaly score screening can reduce inspection burden only when thresholds are governed to control nuisance alarms and periodically recalibrated without absorbing drift into the baseline. The paper concludes with practical guidance for inspection planning, acceptance limit design, baseline management, and traceable release logic aligned with standards oriented requirements for purchased AM parts and critical applications.