STATISTICAL PROCESS

CONTROL (SPC)

PREPARED BY : GOKULKUMAR.K
Training Contents

➢ Basics & History of Quality Management ➢ Sequence for conduct SPC

➢ Statistical Process Control – Definition & Why SPC ? ➢ Types of Control Charts
➢ How Statistical Process Control (SPC) Linked
➢ Points to remember
➢ Causes of Variations
➢ SPC by using Histogram & Control Chart (X-R Chart)
➢ Graph Reading & Why Graph Reading ?

➢ Mean , Standard Deviation & Normal Distribution

Curve

➢ Standard Deviation (S.D) – Definition & How to

Calculate

➢ How to Calculate Probability by using Z-Table & PPM

➢ Skewness
Basics & History of Quality Management

W.E. Deming hired by Ford as a Consultant

➢ 1984 – U.S. Government Designates October as National Quality Month

➢ 1987 – Malcolm Baldridge National Quality Award is Established

➢ 1990’s – Quality Programs Spread to service Industries

Proliferation of Quality Programs : TQM, QFD, Six Sigma, Kaizen, Poka-Yoke, Taguchi Methods,
Benchmarking & FMEA etc.
Basics & History of Quality Management

➢ Pre - Industrial Revolution

➢ 1880’s – Frederick Taylor & “Scientific Management”

➢ 1920’s – Shewhart Introduce Statistical Process Control

➢ Methods based on Continual On-Line Monitoring of Process Variation

➢ Concept of “Common Cause” & “Assignable Cause” Variability.

➢ 1930’s – Dodge & Romig Introduce Acceptance Sampling Methods

➢ 1950’s – W.E. Deming Introduces Statistical Process Control to Japan

➢ 1970’s - U.S. Losing Market Share to Global Competitors

➢ 1980’s – “Quality Revolution” Begins in America

Statistical Process Control (SPC) - Definition

➢ SPC is a collection of Tools that when used together can result in Process Stability & Variability reduction

The Main focus in SPC is,

➢ Quantify the Variation.

➢ Understood the Sources of Variation.

➢ Find ways & means of reducing the variations.

• SPC was developed by Automotive Industry Action Group (AIAG)

• AIAG member’s - GM,Daimler Chrysler corporation & Ford motor company

• Presently we using SPC - AIAG 2nd edition

Statistical Process Control (SPC) - Definition

Why SPC ?

As per IATF 16949 : 2016 - 9.1.1.2 - Identification of Statistical tools for Critical , Special characteristics identified by customer
or Internal

As per IATF 16949 : 2016 - 9.1.1.3 - Application of Statistical Concepts for identified Critical , Special characteristics to be
evaluated

As per APQP - AIAG 2nd edition - 3rd Phase need to identify SPC method for identified Special or Critical Characteristics . 4th
Phase SPC need to conducted & records need to update.
How Statistical Process Control (SPC) Linked

4Th Phase
Advanced Product Quality Planning Production Part Approval Process
(APQP) (PPAP)

1. 3rd Phase - SPC Plan

2. 4th Phase - SPC Reports

Statistical Process Control

(SPC)

Process Capability Index Cpk ( As Per PPAP)

• A significant production run

• A continuous eight-hour production run or a minimum of 300 manufactured parts, whichever comes first. This run shall be
manufactured at the production site using the tooling, gaging, process, materials, and operators from the production environment.
Causes Of Variation

Variation

Common Cause Assignable Cause

Inherent to process No Inherent. Due to Mistakes

Due to many random unidentifiable Due to Specific identifiable causes

sources
Due to 4M change & Environment
Follow predictable pattern condition

➢ Process Improvement Follow Sporadic pattern.

➢Redesign It Can be eliminated.

➢Investment It is also called as Special Cause

Graph Reading
Why Graph Reading ?

➢ Target not need out

➢ Abnormal in rejection % Day to Day

➢ Day to Day random variations need to take corrective actions

➢ Implement actions in further months to reduce Rejection %

Graph Reading

1. Special Cause : It can be eliminated & occurs above the UCL or Below the LCL. Need to identify the possible root cause .
No too much of an investment required compared to common cause

2. Sporadic Cause : Based on 4M Change (It can be adjustable)

3. Common Cause : Due to many random unidentifiable sources .The possible common causes are,

4. Periodic Cause : It may be based on climatic condition variations

5. Increasing / Decreasing : Based on tool adjustment / Tool Wear.

6. Shift Cause : Causes based on 4M Change

7. Chronic Cause : Too much of variations , Issues are chronical & all are known Cause
Mean , Standard Deviation & Normal Distribution Curve
Mean , Standard Deviation & Normal Distribution Curve

1. Bell Shaped
2. Symmetrical
3. Mean,Median & Mode are equal

Mean Median Mode

7,3,4,1,7,6 7,3,4,1,7,6 7,3,4,1,7,6

Median = 7
=(7+3+4+1+7+6)/6 =1,3,4,6,7,7 (Most Common Number)
=28/6 = 4.66 Median = (4+6)/2 = 5
Standard Deviation (S.D) – Definition & How to Calculate

Standard Deviation (S.D) : How much value are deviate from the mean.

Steps need to follow :

➢ Calculate Mean

➢ How the value is deviate from mean (X-Mean)

➢ Calculate (X-Mean)²

➢ Calculate sum of (X-Mean)²

➢ Calculate sum of (X-Mean)² / (N-1)

➢ Calculate Sq. Root of ( sum of (X-Mean)² / (N-1))

Standard Deviation (S.D)

Step -1 Step -2 Step -3 Step -4 & 5

How to Plot Bell Curve

Step -1 Step -2 Step -3 Step -4 & 5

How to Plot Bell Curve

➢ No. Of Readings (N) = 5

➢ Customer Specification = 10 +/- 0.5

➢ Mean = 10.02

➢ S.D (σ) = 0.32

How to Calculate Probability by using Z-Table & PPM

Normal Distribution is also known as the “Z”-Distribution

1. No. Of Readings (N) = 5

2. Customer Specification = 10 +/- 0.5

Find Rejection due to Under size ? Method Used for Sample inspection bsed on
Sampling Standard (IS 2500)
Step-1 : Mean = 10.02

Step-2 : S.D (σ) = 0.32

Step-3 : Z value = X-Mean / Standard Deviation

Z Value = (9.5-10.02) / 0.32 = -1.625

Step-4 : Take values from Z-Table :(-1.625) = 0.0516

= 0.0516*100 = 5.16 %

Step-5 : Calculate PPM = 1-0.0516 = 0.9484 = 94.84 %

Z-Table
Skewness

Skewness is a measure of the asymmetry of a distribution. A distribution is asymmetrical when

its left and right side are not mirror images
Sequence to conduct SPC

1. Identify Customer defined or internally defined Customer Specification.

2. Identify Specification limit

3. Calculate Process Spread

4. Calculate SD & X Bar

5. Calculate Cp (Process Capability)

6. Calculate Cpk (Process Capability Index)

7. Verify Cp Vs Cpk

8. Plot Graphical Representation of Histogram & Bell Curved Chart

Types of Control Charts
Points to remember

1. Cp≥ Cpk - is acceptable

2. Cp≤ Cpk - is not acceptable

3. Cp,Cpk - For Mass Production & PPAP Run

4. Pp,PpK - For Below 300 Nos Parts & PPAP Run

5. Cp - For identify Machine Capability ( Whether this M/c is suitable for this part Specification meet out )

6. Cpk - For Identify Machine + Operator Setting

7. Cp≥ 1.67 - Machine is OK

8. Cpk≥ 1.67 - (Machine + Operator Setting) OK

9. Cp = Cpk - Process is Well Cantered

SPC by using Histogram & Control Chart (X-R Chart)

1.Histogram & Control Chart is one of the 7 QC Tools .

Both tools are help to identify Part to part variation, Range &
Average of Parts .

1.X Bar Chart– is used to identify Mean Shifting

2.R- Chart - is used to identify Variation

 SPC by using Histogram & Control Chart (X-R Chart)

Capability Study
Part Number : XXXX-B20 Customer Name :

Specification : 4.30 +/- 0.10 LSL : 4.2 USL : 4.4 Instrument : Micrometer

Sample Size : 125 Nos Least Count : 0.01 mm Performed By : Date : 9/14/2023

Sub Group
Sub Group Size
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1 4.34 4.33 4.34 4.32 4.32 4.38 4.34 4.38 4.34 4.41 4.32 4.38 4.34 4.38 4.34 4.32 4.38 4.34 4.38 4.34 4.34 4.32 4.38 4.34 4.38
2 4.32 4.34 4.33 4.33 4.34 4.40 4.42 4.40 4.34 4.32 4.34 4.40 4.42 4.40 4.34 4.34 4.40 4.42 4.40 4.34 4.34 4.34 4.40 4.42 4.40
3 4.35 4.43 4.44 4.34 4.41 4.29 4.41 4.31 4.31 4.42 4.41 4.29 4.41 4.31 4.31 4.41 4.29 4.41 4.31 4.31 4.31 4.41 4.29 4.41 4.31
4 4.34 4.33 4.35 4.29 4.32 4.38 4.33 4.38 4.23 4.38 4.32 4.38 4.33 4.38 4.23 4.32 4.38 4.33 4.38 4.23 4.23 4.32 4.38 4.33 4.38
5 4.36 4.31 4.34 4.36 4.36 4.34 4.32 4.29 4.37 4.33 4.36 4.34 4.32 4.29 4.37 4.36 4.34 4.32 4.29 4.37 4.37 4.36 4.34 4.32 4.29
Average - X Bar 4.342 4.348 4.360 4.328 4.350 4.358 4.364 4.352 4.318 4.372 4.350 4.358 4.364 4.352 4.318 4.350 4.358 4.364 4.352 4.318 4.318 4.350 4.358 4.364 4.352

Range - R 0.04 0.12 0.11 0.07 0.09 0.11 0.10 0.11 0.14 0.10 0.09 0.11 0.10 0.11 0.14 0.09 0.11 0.10 0.11 0.14 0.14 0.09 0.11 0.10 0.11

Step-1 : Identify instrument as per instrument selection method

Step -2 : Put Identification Marks on the product & measure the products

Step-3 : Calculate average of X- Bar (sum of 5 Sub group / No.of sub groups)

Step-4 : Calculate Range (Range = Max. observed reading – Min. Observed reading)
SPC by using Histogram & Control Chart (X-R Chart)

Average 4.349 Step-5 : Average of Average - X Bar

Std Dev (S.D ) 0.043 Step-6 : By using Excel (=Std Dev(Select All Values)) R Bar/D2
Min 4.23 Step-7 : By using Excel (=+Min(Selecet All Values))
Max 4.44 Step-8 : By using Excel (=+Max(Selecet All Values))
Range 0.21 Step-9 : Range =Max.Observed Reading - Min. Observed Reading
Interval 11 Step-10 : Interval = Sqrt(Range)
Tolerance 0.2 Step-11 : Interval = USL-LSL
Class Width 0.019 Step-12 : Class Width = Range / Interval
Cp 0.769312125 Step-13 : Cp = (USL - LSL) / (6X S.D)
CpU 0.394503258 Step-14 : CpU = (USL - Average) / (3X S.D)
CpL 1.144120992 Step-15 : CpL = (Average-LSL) / (3X S.D)
Cpk 0.394503258 Step-16 : Cpk = Min(CpU,CpL)
 SPC by using Histogram & Control Chart (X-R Chart)

Step-17 : Determine Class width

(Min observed Reading & sum of (Min observed reading + Class Width )

Step -18 : Identify Cumulative frequency (Ex :4.23~4.25 - 4 Readings)

Step-19 : Remember to don’t take maximum reading again to take another band
(Ex.1st Band 4.23~4.25 & 2nd Band 4.25~4.27)
SPC by using Histogram & Control Chart (X-R Chart)

# Class Width Frequency Cum Frequency Band Width

0 4.23 4.25 4 4 4.23~4.25

1 4.25 4.27 0 4 4.25~4.27
2 4.27 4.29 0 4 4.27~4.29
3 4.29 4.31 0 4 4.29~4.31
4 4.31 4.33 9 13 4.31~4.33
5 4.33 4.35 24 37 4.33~4.35
6 4.35 4.37 35 72 4.35~4.37
7 4.37 4.39 8 80 4.37~4.39
8 4.39 4.41 21 101 4.39~4.41
9 4.41 4.43 8 109 4.41~4.43
10 4.43 4.45 14 123 4.43~4.45
11 4.45 4.47 2 125 4.45~4.47
125 125
Plot a Distribution Chart (Bell Curve Chart)

Band Width Freq

4.23~4.25 4
4.25~4.27 0
4.27~4.29 0
4.29~4.31 9
4.31~4.33 9
4.33~4.35 24
4.35~4.37 28
4.37~4.39 10
4.39~4.41 17
4.41~4.43 17
4.43~4.45 6
4.45~4.47 1
Calculate & Plot X-Bar & R Chart

Calculate X- Bar Chart

Step-1 : Calculate X Double Bar = Sum of. X Bar / No.of X Bar

Step-2 : Calculate X bar Upper & Lower Control Limits

1. UCL X = X Double Bar + A2 *R Bar
2. LCLX = X Double Bar - A2 *R Bar

Calculate R Chart
Step-1 : Calculate R Bar = Sum. Of R/ No.of R

Step-2 : Calculate R Upper & Lower Control Limits

1. UCL R = D4*R Bar
2. LCLR = D3* R Bar
Calculate & Plot X-Bar & R Chart

Capability Study
Part Number : XXXX-B20 Customer Name :

Specification : 4.30 +/- 0.10 LSL : 4.2 USL : 4.4 Instrument : Micrometer

Sample Size : 125 Nos Least Count : 0.01 mm Performed By : Date : 9/14/2023

Sub Group
Sub Group Size
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1 4.34 4.33 4.34 4.32 4.32 4.38 4.34 4.38 4.34 4.41 4.32 4.38 4.34 4.38 4.34 4.32 4.38 4.34 4.38 4.34 4.34 4.32 4.38 4.34 4.38
2 4.32 4.34 4.33 4.33 4.34 4.40 4.42 4.40 4.34 4.32 4.34 4.40 4.42 4.40 4.34 4.34 4.40 4.42 4.40 4.34 4.34 4.34 4.40 4.42 4.40
3 4.35 4.43 4.44 4.34 4.41 4.29 4.41 4.31 4.31 4.42 4.41 4.29 4.41 4.31 4.31 4.41 4.29 4.41 4.31 4.31 4.31 4.41 4.29 4.41 4.31
4 4.34 4.33 4.35 4.29 4.32 4.38 4.33 4.38 4.23 4.38 4.32 4.38 4.33 4.38 4.23 4.32 4.38 4.33 4.38 4.23 4.23 4.32 4.38 4.33 4.38
5 4.36 4.31 4.34 4.36 4.36 4.34 4.32 4.29 4.37 4.33 4.36 4.34 4.32 4.29 4.37 4.36 4.34 4.32 4.29 4.37 4.37 4.36 4.34 4.32 4.29
Average - X Bar 4.342 4.348 4.360 4.328 4.350 4.358 4.364 4.352 4.318 4.372 4.350 4.358 4.364 4.352 4.318 4.350 4.358 4.364 4.352 4.318 4.318 4.350 4.358 4.364 4.352

Range - R 0.04 0.12 0.11 0.07 0.09 0.11 0.10 0.11 0.14 0.10 0.09 0.11 0.10 0.11 0.14 0.09 0.11 0.10 0.11 0.14 0.14 0.09 0.11 0.10 0.11

X Double Bar 4.349

R Bar 0.106

Average Chart Calculation

UCL 4.410

LCL 4.288

Range Chart Calculation

UCL 0.223

LCL 0.000
Calculate & Plot X-Bar & R Chart
Calculate & Plot X-Bar & R Chart
THANK YOU