Statistical Process Control in Textile Manufacturing
topic
Statistical process control (SPC) applies control charts, process capability indices, and sampling theory to textile manufacturing data — enabling production teams to distinguish common cause variation (inherent process randomness) from special cause variation (assignable defect sources) and intervene only when statistically justified, preventing both over-adjustment (adding variation through unnecessary process changes) and under-adjustment (allowing genuine process shifts to generate defects). Shewhart control charts in spinning: X̄-R chart for yarn count control — sample 5 bobbins per spindle set per hour, measure count (tex), plot X̄ (mean) and R (range) → Upper Control Limit UCL = X̄ + A₂ × R̄ (A₂ = 0.577 for n=5), LCL = X̄ − A₂ × R̄ → UCL at Ne 30 mean = 19.68 tex, σ = 0.25 tex: UCL = 19.68 + 3×0.25 = 20.43 tex = Ne 28.9; LCL = 18.93 tex = Ne 31.2 — point outside control limits triggers investigation of autoleveller, drafting roll eccentricity, or roving creel tension. Process capability index Cp and Cpk for dyehouse pH control: dyebath pH target 10.5 ±0.3 (specification limits 10.2–10.8); process standard deviation 0.08 pH units → Cp = (USL − LSL) / 6σ = 0.6 / 0.48 = 1.25 (capable); Cpk = min[(USL − X̄)/3σ, (X̄ − LSL)/3σ] = min[0.3/0.24, 0.3/0.24] = 1.25 (centred, fully capable — target Cpk ≥ 1.33 for Six Sigma quality). Acceptance Quality Limit (AQL) sampling plans (ISO 2859-1) in textile inspection: AQL 2.5 (general inspection level II) for fabric defects — lot size 5,000 m: sample 200 m, Ac (acceptance number) 10, Re (rejection number) 11 — if >10 defects found in 200 m sample, lot rejected. AQL 1.0 for critical defects (safety-related fabric failures). Inline SPC versus end-of-line inspection: inline sensor (USTER Tester 6, evenness CV% measured every bobbin) generates 50,000 data points/shift versus end-of-line sampling 20 bobbins/shift — inline SPC detects process drift within 15 minutes versus 4-hour detection lag with end-of-line sampling, preventing 95% of defective production versus 40% with end-of-line inspection. Economic benefit of SPC in fabric weaving: loom stop frequency reduction (warp breaks per 100 loom-hours reduced from 12 to 8 through SPC-guided process control) → loom efficiency improvement from 88% to 91.5% → additional production 3.5% × 50 looms × $500/loom/day = $875/day = $320,000/year.
Role
Statistical process control is the quantitative methodology that converts quality control from reactive inspection to proactive process management in textile manufacturing — with inline SPC detecting process drift 15× faster than end-of-line sampling and preventing 95% versus 40% of defective production, the $320,000/year loom efficiency improvement from SPC-guided warp break reduction demonstrates that SPC investment is the foundation of systematic quality cost reduction rather than an overhead inspection cost.