SPC Basics

Abstract

Statistical Process Control (SPC) is a quality management tool that uses statistical methods to monitor and control production processes. Its primary purpose is to ensure processes operate under controlled conditions and reduce variation to improve product quality.

Publication
SPC, 1(1)

Core Concepts

  1. Process Variation:
  • Common Causes: Natural variations inherent in the process, usually stable and predictable.
  • Special Causes: Variations caused by specific events or circumstances, typically abnormal and requiring investigation.
  1. Specification Limits:
  • Upper Specification Limit (USL): Maximum allowable value.
  • Lower Specification Limit (LSL): Minimum allowable value.
  1. Control Limits:
  • Upper Control Limit (UCL): Predetermined upper limit in control charts, typically mean plus 3 standard deviations.
  • Lower Control Limit (LCL): Predetermined lower limit in control charts, typically mean minus 3 standard deviations.

In Statistical Process Control (SPC), Cp, Cpk, Pp, and Ppk are four key process capability indices used to evaluate manufacturing process performance.

Cp & Pp: Measure process capability without considering centering. Higher values indicate better capability.

Cpk & Ppk: Measure process performance considering both centering and variation. Higher values indicate better performance.

Generally, Cpk is larger than Ppk because:

  • Cpk reflects short-term capability using controlled subgroup data
  • Ppk reflects long-term performance using overall process data that includes more variation sources

Shift vs Variation

Shift: Difference between process mean and target value.
Variation: Dispersion of process outputs around the mean (measured by σ).

Cp (Process Capability Index) - Short Term


Measures whether process variation fits within specification limits:

Cp Formula

Criteria:

  • Cp > 1.33: Good capability
  • Cp = 1.33: Minimum acceptable
  • Cp < 1.33: Requires improvement

Cpk (Process Capability Index) - Short Term


Considers both variation and centering:

Cpk Formula

Criteria:

  • Cpk > 1.33: Good performance
  • Cpk = 1.33: Minimum acceptable
  • Cpk < 1.33: Requires improvement

Pp (Process Performance Index) - Long Term


Similar to Cp but uses overall process data:

Pp Formula

Criteria:

  • Pp > 1.33: Good performance
  • Pp = 1.33: Minimum acceptable
  • Pp < 1.33: Requires improvement

Ppk (Process Performance Index) - Long Term


Considers both variation and centering for overall process:

Ppk Formula

Criteria:

  • Ppk > 1.33: Good performance
  • Ppk = 1.33: Minimum acceptable
  • Ppk < 1.33: Requires improvement

Control Charts


Common Types:

  1. X-Bar & R Chart:
  • X-Bar: Monitors sample means
  • R Chart: Monitors sample ranges
  1. X-Bar & S Chart:
  • X-Bar: Monitors sample means
  • S Chart: Monitors sample standard deviations
  1. I-MR Chart:
  • Individual Chart: Tracks single measurements
  • Moving Range Chart: Tracks differences between consecutive points
  1. P Chart: Monitors proportion defective
  2. NP Chart: Tracks number of defectives in constant sample sizes
  3. C Chart: Counts number of defects per unit
  4. U Chart: Tracks defects per unit with variable sample sizes

Six Sigma
Quality improvement methodology aiming for 3.4 defects per million opportunities (DPMO). Key metrics:

  • Sigma Level: Process capability measurement
  • DPMO: Defects Per Million Opportunities
Copyright Statement: This article is created by [flexiblebi], copyright belongs to [flexiblebi]. Please include original source link when reposting: [https://flexiblebi.com/publication/spc-basic]
Professor Liu
Professor Liu
Mechanical Engineering Expert

An authority in the field of mechanical engineering, with research interests spanning mechanical design, automation systems, and intelligent manufacturing.