FAQ of SPC

Frequently-Asked Questions for Control Chart Implementation

• Selection of Sample size
• Selection of Sampling frequency
• If we have UCLs and LCLs calculated from same process with different Sample Size Scheme, n=2, n=3, n=4, …, will they be the same values ? Yes, approximately.
• Statistical Process Control
• Characteristics to be measured and plotted on Control Chart could be both Product Characteristic or Process Characteristic (so called Process Parameter) and the benefit you can gain is still the same, as long as you focus on REAL-TIME Monitoring.
• Different perspective of judging Common or Special Cause Variation of the same group of product. Example is acceptance of porosity level of casting product from 2 different technologies. One may perceive that the defect is acceptable as common cause variation, while the other which have better technology may cry out for presence of special cause variation.
• Some special cause variation patterns may lead to opportunities for improvement such as hugging at +/- 1 sigma zone.
• Samples within the same subgroup are not necessarily taken from the product which were consecutively produced. But the essence is that the Variation within any same subgroup should represent "Common Cause Variation of interested process".
• Samples between subgroups are taken from products which were produced in different chronological time. Selection of sampling frequency should depend on combination of these following factors :

1. frequency of changes of main 5 basic factors :

1. Man
2. Machine and equipment
3. Method
4. Material
5. Mother nature (Environment)

Within subgroup variation V.S. Between Subgroup Variation

X-bar chart	R chart
Tracking Between subgroups variation	Tracking Within subgroups variation
Detect Special Cause Variation	Represent Common Cause Variation

Two Definition of Cpk

Product Capability = Total Process Variation,

Cpk = [min (abs(USL-X_doublebar), abs(X_doublebar-LSL)] / 3 Sigma_x

Sigma_x = Standard Deviation obtained from all 125 samples(from all 25 subgroups)

Process Capability = Inherent Process Variation,

Cpk = [min (abs(USL-X_doublebar), abs(X_doublebar-LSL)] / 3 S_x

S_x = Standard Deviation obtained from R_bar / d₂

Selection of X_bar, R chart in stead of X_bar, S chart

SPC Implementation Plan and Q&A

Selection of Characteristic or Parameter for Plotting Control Chart

1. Must be Variable Data, if you want to use X_Bar, R Chart

2. Special Characteristics (defined by Customer or APQP team)

3. Tightest Tolerance Characteristics

How to formulate all population -> Histogram or Dot plot

-> sample -> Histogram or Dot plot

è Subgroup average -> X_bar, R chart

Variable V.S. Attribute data

Example of Terms used to describe and frequency in “Control Plan”

Comparison table for selection of proper sample size and sampling frequency

Sampling Patterns	Units produced between interval of each sample in same subgroup (if not contiguous)	Units produced within interval of each complete subgroup	Units produced between interval of each complete subgroup	Total time for all 25 subgroups	Total units covering 25 subgroups
2 pcs / 24 min
5 pcs / hour
10 pcs / 2 hours

And other consideration such as :

Number of streams, cycles,

The term unit may be replaced by passing time in second or minute for bulk material.