Cp & Cpk

(Health of a Process)
In Order to understand process capability(Cp,Cpk) we have to understand
Process, Input, Output, Process control and some statistics.
Contents
1. Process 2.Input 3. Output 4. Process Control

5.Mean 6.Sigma 7. Cp 8. Cpk 9. Sigma level

Sunjjoy Uvach
1
 Cp & Cpk
(Health of a Process)
In Order to understand process capability(Cp,Cpk) we have to understand
Process, Input, Output, Process control and some statistics.
Contents
1. Process 2.Input 3. Output 4. Process Control

5.Mean 6.Sigma 7. Cp 8. Cpk 9. Sigma level

Sunjjoy Uvach
2
 Cp & Cpk
(Health of a Process)
In Order to understand process capability(Cp,Cpk) we have to understand
Process, Input, Output, Process control and some statistics.
Contents
1. Process 2.Input 3. Output 4. Process Control

5.Mean 6.Sigma 7. Cp 8. Cpk 9. Sigma level

Sunjjoy Uvach
3
cIC

1.Process

Process is something which can be measured, controlled

and improved

To Produce something , Process is required

Every Process has an input and output

INPUT Process OUTPUT

4
Sunjjoy Uvach
cIC

2.INPUT
INPUT Process OUTPUT

To Produce something Input is required

Outcome of Process depends on Input quality

Example: To Fill the Tank with water we require Electricity to drive the pump

Here Electricity is the Input

Stable Voltage.
Quality of Electricity is important i.e., _____________

5
Sunjjoy Uvach
cIC
3. OUTPUT INPUT Process OUTPUT

It is the requirement of the product from the process

Output quality depends on Process and Input.

Example: To Fill the Tank quickly with water

a. Process, Pump should work as per rating &

b. Input, Electricity (Voltage) should be stable.

6
Sunjjoy Uvach
cIC

4.Process Control:
The Output of every process should be controlled within specified
limits given by the customer.

Example: If customer specification for thickness is 10±0.5mm

In this case LSL (Lower specification limit) is 9.5 &
USL (Upper specification limit) is 10.5 mm

In this case we have to control process within LSL 9.5 and USL
10.5

If any thing goes below of above these values will be termed as

rejection because it is not acceptable to the customer.
cIC ;

5. Mean (x̄) :
It is the average of the data collected and denoted by the x-bar symbol,
 x̄, is used in statistics to represent the mean of a set of data.

This is the arithmetic mean, which is what most people think of when they say
"average" – the sum of all the measurements divided by the total number of
measurements.

Example: For a set of measurements x1,x2,x3……………xn, total measurements=N

Mean (Average) x̄= (x1+x2+x3………….+xn)/N
Average age in India is 30 years.
This Mean is the value around which most of the measurements exists.
8
Sunjjoy Uvach
cIC
6. Standard Deviation (σ)
In statistics, the standard deviation (SD, also represented by the
Greek letter sigma σ or the Latin letter s is a measure that is used to
quantify the amount of variation or dispersion of a set of data values.
LSL USL For understanding the process
variations with respect to mean, we
have to study the standard
deviation. If standard deviation is
 x̄ less it means that values are closer
to the mean hence chances of
rejections will be very-very less.
Contd….
Sunjjoy Uvach
cIC
6.Standard Deviation (σ):
For, Sample Data For, Population data
From one batch of production if some 50 Full one batch data is taken.
samples are taken randomly

σ σ

Here x1,x2………..xn are the measurements taken for n nos of sample or full lot of n nos
 cIC
7. Process Capability (Cp)
When a process is running on a continuous basis then one should know its capability to
produce good products conforming to customer requirements.
It is defined by Juran as, “Process capability is the measured, inherent reproducibility of the
product turned out by a process”.
This Cp represents how much attention is to be given to inspection as it adds cost to the
process.
Cp= (USL-LSL)/6σ
Cp & σ are inversely proportional to each other
This means that if σ is more Cp is less & if it is less then Cp is more
Cp is calculated only if behavior of the process is normal.
If Cp<1.33 then 100% inspection is required
If 1.33<Cp<1.67 then sampling inspection can be introduced otherwise one has to do 100%
inspection.
If 1.67<Cp< 2.0 then one should think of increasing production further and inspection can
be minimized further Sunjjoy Uvach
11
 cIC 8. Process Capability (Cpk)

In statistics, a negatively skewed (also known as left-skewed) distribution

is a type of distribution in which more values are concentrated on the
right side (tail) of the distribution graph while the left tail of the
distribution graph is longer.

Skew type depends on the value of Median and Mean

Median: The median of a data set can be found by arranging all the values
from lowest to highest value and picking the one in the middle. If there is
an odd number of data values then the median will be the value in the
middle. If there is an even number of data values the median is the mean
of the two data values in the middle.
cIC

Example: For 1,3,5,7,9 the median is 5

& For 1,3,5,7,9,11 the median is (5+7)/2= 6

If Median< Mean then -ve Skewed

If Median> Mean then +ve skewed
 cIC
8. Process Capability (Cpk)
When a process is running on a continuous basis but it’s output measurements are not
distributed normally i.e. it is skewed from the mean, then Cp does not represent the
correct picture.
In this case we calculate Cpk to understand the process output
Cpk is classified in upper and lower as follows:

Cpu=(USL-X̅ ) / 3σ and

Cpl= (X̅-LSL) / 3σ
Then Cpk will be the lower of the two.

This shows the true process capability as in process environment the data obtained are
skewed from the mean and not uniformly distributed.
14
Sunjjoy Uvach
cIC 8. Process Capability Calculation (Cp,Cpk)
Cp = tolerance
------------------
Total variation ( 6 sigma)
= This indice is called potential process
capability indice

PROCESS CAPABILITY INDEX [ PCI ]

C p = TOL / 6 Sigma

Ex: USL: 50.10, LSL= 50.00

Mean 50.08; Sigma = 0.01, Process is under
Statistical control
C p = 0.1 = 1.67
6 x 0.01
Cp does not describe the actual capability of
process, i.e it does not consider where the
process is set.
cIC 9. Sigma level
Sigma level denotes the rejection PPM (Parts per million) from the said process

Sigma level= 3*Cpk

1 SIGMA(σ)= 697,672

2 SIGMA(σ)= 308,770

3 SIGMA(σ)= 66,811

4 SIGMA(σ)= 6210

5 SIGMA(σ)= 233

6 SIGMA(σ)= 3.4
cIC
10. DIFFERENCE BETWEEN Pp, Ppk &
Cp, Cpk :
• Pp, Ppk • Cp, Cpk
• Index is process performance index • These are process capability index
• This is used during PPAP / short • This is used during ongoing
process capability study process capability study
• Used for stable & chronically • Used only for stable processes
unstable processes
• Captures variation due to both • Captures variation due to only
common causes & special causes common causes.
• Standard deviation = Calculated • Standard deviation = R bar/ d2
considering all individual reading
using calculator or PC
• Customer requirements are: • Customer requirements are :
• Ppk > 1.67 • Cpk > 1.33
 Engineering Engineers
cIC

Questions are Welcome

Thanks for Patience