CONTROL CHART FOR ATTRIBUTES

ATTRIBUTE:

 Attribute is a quality characteristic for which a numerical value is not specified.

 It is measured on a normal scale.

NON – CONFORMITY (DEFECT)

A Quality characteristic that does not meet certain prescribed

standards(Specifications).

NON – CONFORMING ITEM (DEFECTIVE):

A product with one or more non – conformities, unable to meet the intended
standards and is unable to function as required.
ADVANTAGES OF ATTRIBUTE CHARTS:

(i) Provides overall quality information at a fraction of cost.

H
W
L

Here, H is the height

L is the length

W is the width

For variable = 3 chart

For attribute = 1 chart

ADVANTAGES OF ATTRIBUTE CHARTS:

(ii) Used to summarise information about several components that make up a product.

(iii) Encountered that all levels of an organisation (i.e.) company, plant, department, work center
and machine (or) operator level.

(iv) Assists in going from general to a more focused level.

DISADVANTAGES OF ATTRIBUTE CHARTS:

(v) Does not state the degree to which specifications are met or not.

(vi) Does not detect a lack of control until the process parameters are sufficiently changed such
that some non – conforming items are produced.

(vii) Requires large sample sizes than variable charts to ensure adequate protection against a
certain level of process changes.

PRELIMINARY DECISIONS:

No historical information, choice of sample size, summary measures, information

about the output at plant – level.
CHART FOR PROPORTION NON – CONFORMING (P - CHART)
 Based on binomial distribution.

 samples must be independent for a given sample.

The proportion non – conforming

= P^ = x/n

= No. of non – conforming items in the sample /Sample size

 Versatile control charts.

 Used to control the acceptability of a single quality characteristic, a group of quality

characteristic of the same type or on the same part or an entire product.

 Provides a fair indication of the general state of a process.

 Good tool for relating information about the average quality level to top management.

 Measure a performance of top management.

 Provides a source of information for improving product quality.

 Identify the circumstances for which X̄ and R chart should used.

CONSTRUCTION AND INTERPRETATION

(1) Select the objective:

No. of potential quality characteristic, the No. of products produced, Cost and time
required for inspection.

(2) Determine the sample size and sampling interval.

(3) obtain the data and record on an appropriate form.

(4) Calculate the center line and trial control limits.

No standard or target value specified,

P^ = x/n

Center line p = CLP = P̄ = Σ P^i/g = Σ xi/g ; i= 1 to g

` (g is the No. of samples)
True value of P is not known,

UCLP = P̄ + 3√[P̄ (1- P̄ )/n]

LCLP = P̄ - 3√[P̄ (1- P̄ )/n]

CONSTRUCTION AND INTERPRETATION

Standard or target value specified,

Center line p = CLP = Po (Po is the standard or target value)

UCLP = Po + 3√[Po(1- Po)/n]

LCLP = Po - 3√[Po(1- Po)/n]

(v) Calculate the revised control limits.

Revised process average P̄ r is calculated.

The revised center line and control limits

CLP = P̄ r
UCLP = Pr + 3√[Pr(1- Pr)/n]

LCLP = Pr - 3√[Pr(1- Pr)/n]

CONSTRUCTION AND INTERPRETATION

(vi) Implement the chart.

Variable sample size

Control limits for individual samples (No standard is given)

UCLP = P̄ + 3√[P̄ (1- P̄ )/ni]

LCLP = P̄ - 3√[P̄ (1- P̄ )/ni]

Control limits based on average sample size

The average sample size n̄ = Σ ni/g , g is the No. of samples.

Control limits are then given by,

UCLP = P̄ + 3√[P̄ (1- P̄ )/n̄ ]

LCLP = P̄ - 3√[P̄ (1- P̄ )/n̄ ]

Standardised control chart,

E(P^) = P

σP^ = √[p(1-p)/n]
Standardised value of the proportion non – conforming for the i^th sample

Z = (P^I - P̄)/√[P̄(1- P̄)/ni]

CHART FOR NUMBER OF NON – CONFORMING ITEMS (np - CHART)

 Counting the No. of non – conforming items in the sample.

 Binomial distribution is used.

 If the sample size changes, the center line and control limits change as well.

 np chart should not be used when sample size varies.

No standard given (np – chart)

Center line = CLnp = Σ xi/g , i= 1 to g

= np̄

xi is the number non conforming for the i^th sample

g is the No. of sample

n is the sample size

p̄ is the sample average proportion non – conforming

CHART FOR NUMBER OF NON – CONFORMING ITEMS (np - CHART)

 the standard deviation of the non – conforming = σnp = √[nP̄(1- P̄)]

 Control limits for an np chart are

UCLnp = nP̄+3√[nP̄(1- P̄)]

LCLnp = nP̄-3√[nP̄(1- P̄)]

Standard given , npo = Specified standard for the number of non – conforming items.

Center line = CLnp = npo

UCLnp = npo +3√[npo(1- po)]

LCLnp = npo -3√[npo(1- po)]

CHART FOR THE NUMBER OF NON – CONFORMITIES (c - CHART)

 C – chart is used to track the total No. of non – conformities in samples of constant size.

 Sample size varies, a U – chart is used to track the No. of non – conformities per unit.

 Follows a poisson distribution

No standard given
The average No. of non – conformities per sample unitis found from the sample
observations and is denoted by c̄.

Center line = CLc = c̄

UCLc = c̄ + √c̄

LCLc = c̄ - √c̄

Standard given
CLco= co ; co is the specified goal for the No. of non – conformities per
sample unit

UCLco = co + √co

LCLco = co - √co
CHART FOR NUMBER OF NON – CONFORMITIES PER UNIT (u - CHART)

 For sample size variation.

 Constructed to monitor the number of non – conformities per unit.

Variable sample size and no specified standard

the No. of non – conformites / unit for the i^th sample

 Ui = Ci/ni

Ci : the No. of non – conformities in the i^th sample

ni : the size of the i^th sample

The average No. of non – conformities / unit (Ū)

Ū = Σ Ci / Σ ni ; i= 1 to g

The control limits are given by,

UCLU = Ū + 3 √(Ū/ni)

LCLU = Ū - 3 √(Ū/ni)
CHART FOR DEMERITS PER UNIT (U - CHART)

 Quality rating system, which rates demerits per unit is called the U – chart.

 Helpful in service applications.

Classification of Non – conformities:

It is classified in 4 categories:
(1) Class 1 defects (Very serious):

Defects that lead directly to severe injury or to catastrophic economic loss.

(2) Class 2 defects (Serious):

Defects that lead to significant injury or to significant economic loss.

(3) Class 3 defects (Major):

Defects that can cause major problems with normal use of a product or service
rendered.

(4) Class 4 defects (Minor):

Defects that can cause minor problems with normal use of a product / service
rendered.
Construction of U – chart

Assume 4 categories of non – conformities

Sample size n.

Total No. of non – conformities in a sample C1, C2, C3 and C4.

Weights assigned to each category W1, W2, W3 and W4.

Non – conformities in each category are independent of defects in other categories.

Follows poisson distribution.

The total no of demerits is given by,

D = W1C1 + W2C2 + W3C3 + W4C4
the demerits / unit for the sample

U= D/n = (W1C1 + W2C2 + W3C3 + W4C4)/n

Ū=W1 Ū1 + W2Ū2 + W3Ū3 + W4Ū4

Where, Ū1, Ū2, Ū3, and Ū4 are the average No. of non – conformities/unit in their respective
classes
The estimated Standard deviation of U,

σ^u = √[{w1^2Ū1 + w2^2Ū2 + w3^2Ū3 + w4^2Ū4}/n]

Control limits for U – Chart

UCLu = Ū + 3σ^u

LCL L = Ū - 3σˆu