SPC - Statistical process control
Download as PPT, PDF6 likes5,894 views
The document provides an overview of statistical process control (SPC) techniques. It discusses key SPC concepts like process characteristics, sources of variation, control charts, and metrics like Cp, Cpk, Pp, and Ppk. The document is intended as an internal presentation on SPC fundamentals and how to select and interpret different control chart types to monitor variable or attribute process data. The goal is to help understand whether a process is stable or not using statistical analysis of sample measurements.
SPC - Statistical process control
- 1. SPC For Internal Use Only Presentation By Senthilkumar RM
- 2. Today Learning's !! 1.Product Characteristics 2.Source of Process 3.Process Types 4.Variation & Its Causes 5.Tolerance 6.Standard Deviation 7.Interpretation of Stdev with Distribution 8.Calculating Std deviation 9. SPC ,Control Chart Selection & It interprets 10.Cp,CpK,Pp,Ppk For Internal Use Only Presentation By Senthilkumar RM
- 3. Product Characteristics • 2 Types of Product Characteristics • Variable • Attribute • Variable – A Characteristics measured in Physical units eg : Millimeter, Volts • Attributes – A characteristics measured by comparison eg : Good and Bad For Internal Use Only Presentation By Senthilkumar RM
- 4. Source of Process For Internal Use Only Presentation By Senthilkumar RM
- 5. Process Types All Process have • Natural Variability (due to Common cause) • Un-Natural Variability (due to Special cause) • Stable Process – Variation in Output is by only common cause • Unstable Process – Variation in Output is by both common & Special cause For Internal Use Only Presentation By Senthilkumar RM
- 6. Variation & Its causes Variation : Inevitable changes in output over time Two causes of Variation • Common Cause : Implemented in process due to design of process • Special Cause : Not Present in process arises in certain circumstances Answer : • Poor design ? • Broken part ? For Internal Use Only Presentation By Senthilkumar RM
- 7. Tolerance Tolerance : Allowed Maximum variation in dimension Otherwise called as “Engineering Tolerance” For Internal Use Only Presentation By Senthilkumar RM
- 8. Quick recap 1. Two types of Product characteristics? 2. 5 source of Process? 3. Two types of Process? 4. Two Cause of variation ? 5. What is Engineering Tolerance ? How Process speaks ? What is Within control and Out of control in a process? For Internal Use Only Presentation By Senthilkumar RM
- 9. Standard Deviation (σ) Standard Deviation is a measure of Variability . It is denoted by Greek letter σ . In Other words Data dispersion around mean What is Population and Sample? For Internal Use Only Presentation By Senthilkumar RM
- 10. Distribution & Standard Deviation and 3σ Level For Internal Use Only Presentation By Senthilkumar RM
- 11. Practical example to understand Distribution Now, we can create an example of our own and understand 2. What is Observation X 3. What is Mean (X bar) 4. How to calculate std deviation? 5. What it interprets in distribution? 6. Histogram or Frequency Chart? Explanation 7. How bell curve is derived? Explanation 8. What is 6 σ ? What it tells? % derivation? For Internal Use Only Presentation By Senthilkumar RM
- 12. What is SPC SPC : Statistical Process control • Statistical tool • Inspects random sample from a process • To determine whether the process is working properly or not • Forms Charts with data (Control charts) and Process speaks • Developed by Walter Schewart For Internal Use Only Presentation By Senthilkumar RM
- 13. Why We need to care about SPC ? Suppose Supplier A parts are with in Spec Supplier B parts are within Spec, But DON’T FIT TOGETHER WHY? For Internal Use Only Presentation By Senthilkumar RM
- 14. Why We need charts Data Charts Information Tools For Internal Use Only Presentation By Senthilkumar RM
- 15. Control charts It is a Chart, named as control charts because it interprets to Us whether the process is in control or not Remember, Two Types of data • Variable data • Attribute data Based on data, We need to select the control charts !! There should be some common factor also Sample size For Internal Use Only Presentation By Senthilkumar RM
- 16. Selection of Control charts Based on Sample size & Data Type – Selection of chart done • Variable data • Attribute data For Internal Use Only Presentation By Senthilkumar RM
- 17. I – IR- MR Charts Sample size = 1 What : “I” stands for Individual; R Stands for Range; MR stands for Moving Range When : Situation like data are limited such as production low, testing or Homogeneous Batches (Heat Treatment) How : By Plotting Individual Measurements and Moving range in separate graph and adding Center line For Internal Use Only Presentation By Senthilkumar RM
- 18. X bar – X bar R – X Bar S chart Sample size = 3 ~5 What : “X Bar” stands for average; R Stands for Range; When : Situation like data are limited such as production High, How : By Plotting average Measurements and range in separate graph and adding Center line with spec (Sub Group-->>>>>) For Internal Use Only Presentation By Senthilkumar RM
- 19. Attribute Charts Attributes data – are “ Count data” where each data is classified in one of two categories eg., Good or Bad ; Pass or Fail Defects – is “ Non conformity ” eg., No.of defects in that unit / board / piece Defectives– is “ Non conformance ” eg., That unit / board / piece which has one or more defects is called defectives What are Defects in item Which of the following is the source of Attribute data’s Cap Loose 1. Cap loose (one of defect) 2. Label wrong •Testing – Why ? 3. Breakage •Productivity – Why ? Label •Caliper measurement – Why ? wrong What is Defective here •Ring gage – Why ? Breakage •Answer paper– Why ? WATER BOTTLE (unit) For Internal Use Only Presentation By Senthilkumar RM 19
- 20. Chart Types For Internal Use Only Presentation By Senthilkumar RM
- 21. Selection of Chart : Attribute See Sample size first Constant Sample size – C chart or nP chart Variable Sample size – P chart or U chart See type of data Defects – C chart or nP chart Defectives – P chart or U chart For Internal Use Only Presentation By Senthilkumar RM
- 22. Now Tell For Internal Use Only Presentation By Senthilkumar RM
- 23. Short Term (Sst) & Long Term (Slt) Standard Deviation For Internal Use Only Presentation By Senthilkumar RM
- 24. Cp,Cpk,Pp,Ppk Cp – Process Capability – • Ability of a production process to meet desired spec • Common Cause capability • Short –Term Pp – Process Performance – • Ability of a production Process to meet desired spec • Includes special cause • Long – Term Cpk, Ppk are indexes – Numbers of Cp, Pp For Internal Use Only Presentation By Senthilkumar RM
- 25. Calculate Cp, Cpk Cp = Engineering Tolerance / 6 σ Engineering Tolerance = USL – LSL Engineering Tolerance with Std Deviation = USL- Avg/ 3 σ = Avg – LSL/ 3 σ Cp = Engineering Tolerance / 6 * Sst – (see next page) (Question here) Cpk = Smaller of Engineering Tolerance with Std Deviation (Question here) For Internal Use Only Presentation By Senthilkumar RM
- 26. For Internal Use Only Presentation By Senthilkumar RM
- 27. Cp,Cpk,Pp,Ppk For Internal Use Only Presentation By Senthilkumar RM
- 28. Data for calculating Cp.. Illustration Observations = Customer Spec = 60 +- 5 6. n= 7. Avg = 3.Std dev = 4.Eng Tolerance = 5.Cp = 6.Cpk = Lets do this XL For Internal Use Only Presentation By Senthilkumar RM
- 29. Higher the better For Internal Use Only Presentation By Senthilkumar RM