WEI_ZHENGTAI_SPC
- 1. Statistical Process Control (SPC) By WEI ZHENGTAI School of Electrical and Electronics Engineering(Power), Nanyang Technological University Diploma in Mechatronics Engineering with MERIT(Wafer Fabrication), Nanyang Polytechnic
- 2. Contents Why SPC SPC-4M+1E Terminology Control Charts Process Improvement Tool: 6σ How to Derive Control Limits Reference
- 3. Why SPC Maintain good product quality Control the production costs(rejection costs included) Provide clear objectives and reduce workloads Easier to detect machine fault and helps in condition-based maintenance Production personnel will gain awareness Gain good business reputation among the customers Return to Content
- 4. SPC-4M + 1E Man Machine Method Material Environment Return to Content
- 5. Terminology Max, Min, Mean Range=𝑋 𝑚𝑎𝑥-𝑋 𝑚𝑖𝑛 MR=𝑋n+i-𝑋 𝑛 Standard Deviation Variance= Return to Content
- 6. Terminology (Cont’d) CL(Control Limit),UCL(Upper Control Limit) and LCL(Lower Control Limit) 𝑋 chart: UCL= 𝑋+3σ= 𝑋+𝐴2 𝑅 LCL= 𝑋-3σ= 𝑋-𝐴2 𝑅 (𝐴2=defined factors used in calculating the control limits) CL= 𝑋 R chart: UCL= 𝑅+3σ=𝐷4 𝑅 LCL= 𝑅–3σ=𝐷3 𝑅 (𝐷4 and 𝐷3 is defined factors used in calculating the control limits) CL= 𝑅 Return to Content
- 7. Terminology (Cont’d) SL(Specification Limit), USL(Upper Specification Limit) and LSL(Lower Specification Limit) PC(Process Capability)=6σ Cp(Capability of Potential Process),Cpk,Cpl(Z(L)),Cpu(Z(U)),Cmk(Machine) and Cr - Cp= 𝑈𝑆𝐿−𝐿𝑆𝐿 6𝜎 If 6σ<specification tolerances => Cp > 1 If 6σ=specification tolerances => Cp = 1 If 6σ>specification tolerances => Cp < 1 - Cmk= 𝑈𝑆𝐿−𝐿𝑆𝐿 8𝜎 - Cr=𝐶𝑝−1 - Cpl= 𝑥−𝐿𝑆𝐿 3𝜎 - Cpu= 𝑈𝑆𝐿− 𝑥 3𝜎 - Cpk=Min(Cpl,Cpu)=Cp(1-K), whereK = |𝑀− 𝑥| (𝑈𝑆𝐿−𝐿𝑆𝐿)/2 , 𝑀 = (𝑈𝑆𝐿 + 𝐿𝑆𝐿)/2 Return to Content
- 8. Control Charts Variable Control Charts - Measurement is critical - Precision required - Accurate test devices - 1 characteristics - Example:Xbar-R, Xbar-S, CuSum charts Attribute Control Charts - Measurement is not possible - Measurement is time consuming - > 1 characteristics -Example: np, c and u charts Return to Content
- 9. Control Charts (Cont’d) Xbar-R charts (average-range) - Average : variability between samples – Range: variability within samples np Control Chart - Determine the defective items produced by a process - Constant sample size - Steps 1.Gather data : Determine sample size (n) Determine sampling frequency or subgroup, (k) Determine total no. of samples (n x k) Record the no. of non-conforming units for each sample group. 2.Calculate process average number of non-conforming p 3.Calculate the Control limits 4.Plot the np chart. Return to Content
- 10. Control Charts (Cont’d) Return to Content np Control Chart Calculation
- 11. Control Charts (Cont’d) Return to Content p Control Chart - Determines the fraction or percentage of defective, whereas the np control chart determines the number of defective - When the number of samples per subgroup is constant or when the number of samples per subgroup varies - Preferred where more people seemed to be able to conceptualized as the data are in terms of percentage defective.
- 12. Control Charts (Cont’d) Return to Content p Control Chart Calculation
- 13. Process Improvement Tool: 6σ Return to Content Introduced by Engineer Bill Smith in 1986, accepted by most of the large companies Keep improving the product and make its behaviour stable Provide important data as a reference of important strategy Methods(DMAIC for business process and DMADV/RDMADV/DFSS for manufacturing and design) - R:Recognition - D:Define - M:Measure - A:analyse - D:Design - I:Improve - V:Verify - C:Control DFSS: Design for six Sigma
- 14. Process Improvement Tool: 6σ(Cont’d) σ Level Defects Per Million Opportunities(DPMO) Yield(%) 1 690,000 30.85 2 308,000 69.15 3 66,810 93.32 4 6,210 99.38 5 230 99.977 6 3.4 99.99966 Return to Content
- 15. How to Derive Control Limits Steps ① Identify the characteristics we need to control ② Select the sample size ③ Data collection ④ Select chart type ⑤ Calculation ⑥ Generate the chart ⑦ Cooperate with operators and seek improvement ⑧ If the situation is not improved, analyse the problem and repeat the previous steps Return to Content
- 16. Reference Course Notes: Topic 6, EGB205 Quality Assurance, Diploma in Mechatronics Engineering, School of Engineering, Nanyang Polytechnic Internship (Globalfoundries) Google Open Search: Six Sigma Return to Content