+ All Categories
Home > Documents > Module 5 Spc

Module 5 Spc

Date post: 07-Dec-2015
Category:
Upload: walter-white
View: 16 times
Download: 0 times
Share this document with a friend
Description:
this is module 5 of spc
Popular Tags:
116
STATISTICAL PROCESS CONTROL
Transcript
Page 1: Module 5 Spc

STATISTICAL PROCESS CONTROL

Page 2: Module 5 Spc

� Statistics is defined as the science that deals with the

� Collection,

� Tabulation,

� Analysis,

� Interpretation,

� Presentation of

quantitative data.

� SPC is one of the best technical tool for improving product and

service quality.

8/11/20152

Page 3: Module 5 Spc

� SPC is a control system which uses statistical techniques for

knowing, all the time, changes in the process.

� It is an effective method

� Helps in preventing defects

� Helps continuous quality improvement.

8/11/20153

Page 4: Module 5 Spc

� Statistical Process Control

Statistical:

� Statistics are tools used to make predictions on performance.

8/11/20154

Page 5: Module 5 Spc

� Statistical Process Control

� Process:

� The process involves

� people,

� machines,

� materials,

� methods,

� management

� and environment

working together to produce an output, such as an end product.

PROCESSPeople

EquipmentMethod

EnvironmentMaterialsProcedures

8/11/20155

Page 6: Module 5 Spc

The Process

People Machines Material

Management Methods Environment

Output

8/11/20156

Page 7: Module 5 Spc

Statistical Process Control

� Control:

� Controlling a process is

� guiding it

� comparing actual performance against a target.

� Then identifying when and what corrective

action is necessary to achieve the target.

8/11/20157

Page 8: Module 5 Spc

� S.P.C. is statistical analysis of

� the predictability

� and capacity of a process

� to give a uniform product.

8/11/20158

Page 9: Module 5 Spc

The Aim of S.P.C. - Detection Strategy

� This focuses on identification of problems after production, by

100% inspection or by customer complaints.

� It is a historically-based strategy.

8/11/20159

Page 10: Module 5 Spc

Detection Drawbacks:

� Production is already made.

� Customer dissatisfaction.

� Inflated costs - rework; inspection.

� Repetitive problems.

� Neglected improvements.

8/11/201510

Page 11: Module 5 Spc

The Aim of S.P.C - Prevention Strategy

� Prevention:

� This focuses on in-process production and identification of

problems through analysis of process capability.

� It is a future-orientated strategy.

8/11/201511

Page 12: Module 5 Spc

Prevention Benefits:

� Improved design and process capability.

� Improved manufacturing quality.

� Improved organisation.

� Continuous Improvement.

8/11/201512

Page 13: Module 5 Spc

� S.P.C. as a Prevention Tool

� The S.P.C. has to be looked at as a stage towards completely

preventing defects.

� With stable processes, the cost of inspection and defects are

significantly reduced.

8/11/201513

Page 14: Module 5 Spc

The Benefits of S.P.C.

� Assesses the design intent.

� Achieves a lower cost by providing an early warning

system.

� Monitors performance, preventing defects.

� Provides a common language for discussing process

performance.

8/11/201514

Page 15: Module 5 Spc

� Process Variations

Process Element Variable Examples

Machine………………………….Speed, operating temperature,

feed rate

Tools………………………………..Shape, wear rate

Fixtures…………………………..Dimensional accuracy

Materials…………………………Composition, dimensions

Operator…………………………Choice of set-up, fatigue

Maintenance…………………Lubrication, calibration

Environment…………………Humidity, temperature

8/11/201515

Page 16: Module 5 Spc

Process Variations

� No industrial process or machine is able to produce consecutive

items which are identical in appearance, length, weight,

thickness etc.

� The differences may be large or very small, but they are always

there.

� The differences are known as ‘variation’.

� This is the reason why ‘tolerances’ are used.

8/11/201516

Page 17: Module 5 Spc

Designed Size

10 11 12 13 14 15 16 17 18 19 20

8/11/201517

Page 18: Module 5 Spc

Natural Variation

14.5 14.6 14.7 14.8 14.9 15.0 15.1 15.2 15.3 15.4 15.5

8/11/201518

Page 19: Module 5 Spc

8/11/201519

Process Variability Variations due to:

Natural Causes:• Temperature variation

• Material variation

• Customer differences

• Operator performance

Special Causes:• Machine is breaking

• Untrained operative

• Machine movement

• Process has changed

Must be monitored Early and visible

warning required

Page 20: Module 5 Spc

Stability

� Common causes are the many sources of variation that are

always present.

� A process operates within ‘normal variation’ when each element

varies in a random manner, within expected limits, such that the

variation cannot be blamed on one element.

� When a process is operating with common causes of variation it

is said to be stable.

8/11/201520

Page 21: Module 5 Spc

Process Control

� The process can only be termed ‘under control’ if it gives

predictable results.

� Its variability is stable over a long period of time.

8/11/201521

Page 22: Module 5 Spc

Statistical Process Control Steps

6-22

Produce GoodProvide Service

Stop Process

Yes

No

Assign.Causes?Take Sample

Inspect Sample

Find Out WhyCreate

Control Chart

Start

8/11/2015

Page 23: Module 5 Spc

� Control Chart Decision Tree

� Determine Sample size (n)

� Variable or Attribute Data

� Variable is measured on a continuous scale

� Attribute is occurrences in n observations

� Determine if sample size is constant or changing

8/11/201523

Page 24: Module 5 Spc

Start

X bar , R

X bar, S

IX, Moving

Range

p (fraction defective)

or

n p (number def. Per sample

c (defects per

sample or

u defects per unitu

Control Chart Decision Tree

8/11/201524

Page 25: Module 5 Spc

What does it look like?

o Adding the element of time will help clarify your

understanding of the causes of variation in the processes.

o A run chart is a line graph of data points organized in time

sequence and centered on the median data value.

8/11/201525

Page 26: Module 5 Spc

Individual X charts

How is it done?

� The data must have a normal distribution (bell curve).

� Have 20 or more data points. Fifteen is the absolute minimum.

� List the data points in time order. Determine the range between each of the consecutive data points.

� Find the mean or average of the data point values.

� Calculate the control limits (three standard deviations)

� Set up the scales for your control chart.

� Draw a solid line representing the data mean.

� Draw the upper and lower control limits.

� Plot the data points in time sequence.

8/11/201526

Page 27: Module 5 Spc

Control Charts� Next, look at the upper and lower

control limits. If your process is in

control, 99.73% of all the data

points will be inside those lines.

� The upper and lower control limits

represent three standard deviations

on either side of the mean.

� Divide the distance between the

centerline and the upper control

limit into three equal zones

representing three standard

deviations. 8/11/201527

Page 28: Module 5 Spc

� Search for trends:

� Two out of three consecutive points

are in zone “C”

� Four out of five consecutive points

on the same side of the center line

are on zone “B” or “C”

� Only one of 10 consecutive points

is in zone “A”

8/11/201528

Page 29: Module 5 Spc

�Basic Control Chartsinterpretation rules:

� Specials are any points abovethe UCL or below the LCL

� A Run violation is seven ormore consecutive points aboveor below the center (20-25 plotpoints)

� A trend violation is any upwardor downward movement offive or more consecutive pointsor drifts of seven or morepoints (10-20 plot points)

� A 1-in-20 violation is morethan one point in twentyconsecutive points close to thecenter line

8/11/201529

Page 30: Module 5 Spc

� Attribute Control charts :

To monitor Attribute data (Characteristics that are measured as

either "acceptable" or "not acceptable", thus have only discrete,

binary, or integer values).

� Variable Control charts :

To monitor the Variable data (Characteristics that are measured

on a continuous scale).

Page 31: Module 5 Spc

� Control charts for Variables

� X-bar chart

� It is used to monitor change in mean of a process.

� Variation in the average of the samples.

� R-chart

it shows the consistencyof the process.

Page 32: Module 5 Spc

Control charts

� Control charts for Attributes

� P-chart

� C-Chart

� np-Chart

� U-Chart

Page 33: Module 5 Spc

CONTROL CHARTS FOR VARIABLES

Page 34: Module 5 Spc

� Sources Of Variation

� Types Of Variation

� Types Of Variable Control Charts

� Control Chart Patterns

� Control Chart And Warning Control Limit

� Basic Equations And Example

� Consequences Of Misinterpreting Control Charts

8/11/201534

Page 35: Module 5 Spc

�Control chart is a graph that displays data taken over time andvariations of this data.

�Control charts, also known as Shewhart charts or process-behavior charts

�One of the axioms or truism of manufacturing is that no twoobjects are ever made exactly alike.

�When variations are very small, it may appear that items areidentical; but precision instruments will show difference.

8/11/201535

Page 36: Module 5 Spc

SOURCES OF VARIATION

Mainly there are four sources of variations. They are,

�Processes

�Materials

�Operators

�Miscellaneous factors

8/11/201536

Page 37: Module 5 Spc

TYPES OF VARIATIONS

There are two kinds of variations. They are,

1. Assignable (or special) causes of variations and

2. Chance (or random or common) causes of variations.

8/11/201537

Page 38: Module 5 Spc

1. Assignable causes of variations

�Assignable causes of variations are larger in magnitude andcan be easily traced and detected.

� The prime objective of control chart is detecting assignablecauses of variation by analyzing data(length, dia.. Etc.)

�Actions on the part of both management and workers willreduce the occurrence of assignable causes.

8/11/201538

Page 39: Module 5 Spc

REASONS FOR ASSIGNABLE CAUSES OF VARIATION

� Difference among machines.

� Difference among materials.

� Difference among process.

� Difference in each of these factors overtime.

� Difference in their relationship to one another. 8/11/201539

Page 40: Module 5 Spc

2. CHANCE (OR RANDOM OR COMMON) CAUSES OF VARIATIONS

� Chances causes of variations are inevitable in any process.

� These are difficult to trace and control even under best conditions of

production.

� All occur at random and cannot be avoided.

8/11/201540

Page 41: Module 5 Spc

REASONS FOR CHANCE CAUSES OF VARIATION

� Human variability from one operation cycle to the next

� Minor variations in raw materials

� Fluctuations in working conditions.

� Lack of adequate supervision skills.

8/11/201541

Page 42: Module 5 Spc

TYPES OF VARIABLE CONTROL CHARTS

�X Bar or average chart

� R or Range chart

8/11/201542

Page 43: Module 5 Spc

CONTROL CHART PATTERNS

1. Natural Patterns

� In natural pattern no points fall outside the control limits.

� The majority of points are near the center line.

2. Sudden shifts in Level

� Many changes can bring about a sudden change( or jump) in

pattern on an X bar and R charts

� Changes in process settings like temperature, depth of cut,

new operators, etc. 8/11/201543

Page 44: Module 5 Spc

CONTROL CHART PATTERNS CONT..

3. Graduation shifts in the Level

� Gradual shifts in the level occurs when a process parameter changesgradually over a period of time.

� Such gradual shift in X bar chart may occur because the incoming quality ofraw materials changed over time, change in style of supervision etc.

� Such shift in R chart may occur because of new operator, decrease inworker skill due to fatigue etc.

4. Tending Pattern

� Trend pattern represents changes that steadily increase or decrease.

� Trend pattern in X bar chart may occur because of tool wear, die wear etc8/11/201544

Page 45: Module 5 Spc

CONTROL CHART PATTERNS CONT..

� Trending pattern in R chart may occur because of Gradual improvement in operator still resulting from on-the-job training.

� Decrease in operator skill due to fatigue.

5. Cyclic Patterns

� Cyclic patterns are characterized by a repetitive periodic behavior in the system.

� Cyclic patterns in X bar chart may occur because of rotation of operators, periodic changes in temperature, humidity etc.

� Cyclic patterns in R chart may occur because of operator fatigue, periodic maintenance of equipment's etc.

8/11/201545

Page 46: Module 5 Spc

CONTROL CHART PATTERNS CONT..

6. Wild Patterns

� Bunching ( or groups) are clusters of several observations

that are decidedly different from other points of the plot.

� Such behavior is due to use of new vendor for a short period

of time, use of a machine for a brief period of time etc.

8/11/201546

Page 47: Module 5 Spc

CONTROL CHART AND WARNING LIMITS

8/11/201547

Page 48: Module 5 Spc

Formula

8/11/201548

Page 49: Module 5 Spc

8/11/201549

Page 50: Module 5 Spc

8/11/201550

Page 51: Module 5 Spc

Example problem for X bar and R Chart

8/11/201551

� The goliath Tool company produces slip ring bearings which

look like washers. They fit around shaft such as drive shafts in

machinery or motors. In the production process for a

particular slip ring bearing the employees have taken 10

samples (during a 10 day period) of 5 slip ring bearings. The

individual observations are shown below. Prepare the R chart

and X bar chart.

Page 52: Module 5 Spc

8/11/201552

Sample

No

Observation (slip ring diameters cm)

1 2 3 4 5

1 5.02 5.01 4.94 4.99 4.96

2 5.01 5.03 5.07 4.95 4.96

3 4.99 5 4.93 4.92 4.99

4 5.03 4.91 5.01 4.98 4.89

5 4.95 4.92 5.03 5.05 5.01

6 4.97 5.06 5.06 4.96 5.03

7 5.05 5.01 5.1 4.96 4.99

8 5.09 5.1 5 4.99 5.08

9 5.14 5.1 4.99 5.08 5.09

10 5.01 4.98 5.08 5.07 4.99

Page 53: Module 5 Spc

CONSEQUENCES OF MISINTERPRETING CONTROL CHARTS

� Blaming people for problems that they cannot control.

� Spending time and money looking for problems that do not exist.

� Spending time and money on unnecessary process adjustments.

� Taking action where no action is warranted.

8/11/201553

Page 54: Module 5 Spc

�Our goal is to decrease the variation inherent in a

process/material/operator over time.

�As we improve those factors ( mentioned above), the spread of

the data will continue to decrease.

�Quality improves!!

8/11/201554

Page 55: Module 5 Spc
Page 56: Module 5 Spc

“Process Capability”

is the ability of a process to make a feature within its tolerance.

Page 57: Module 5 Spc

� Process Capability

A critical aspect of statistical quality control is evaluating the

ability of a production process to meet or exceed preset

specifications. This is called process capability.

� Product specifications, often called tolerances, are preset

ranges of acceptable quality characteristics, such as product

dimensions.

Page 58: Module 5 Spc

� For a product to be considered acceptable, its characteristics

must fall within this preset range.

� Otherwise, the product is not acceptable.

� Product specifications, or tolerance limits, are usually

established by design engineers or product design specialists.

Page 59: Module 5 Spc

� Process capability is the ability of the process to meet the

design specifications for a service or product.

� Nominal value is a target for design specifications.

� Tolerance is an allowance above or below the nominal value.

Page 60: Module 5 Spc

20 25 30

Upperspecification

Lowerspecification

Nominalvalue

Process is capable

Process distribution

Page 61: Module 5 Spc

Process is not capable

20 25 30

Upperspecification

Lowerspecification

Nominalvalue

Process distribution

Page 62: Module 5 Spc

� Process capability is measured by the process capability index

( Cp ) .

� which is computed as the ratio of the specification width to

the width of the process variability

R bar Mean of the sample range

d2 is the value taken from statistical table

2d

RX6= widthProcess

Page 63: Module 5 Spc

Cpk = Minimum ofUpper specification – x

x – Lower specification

3σ,

= =

Process Capability Index, Cpk, is an index that measures the

potential for a process to generate defective outputs relative to

either upper or lower specifications.

Process Capability Index, Cpk

We take the minimum of the two ratios because it gives the worst-

case situation.

Page 64: Module 5 Spc

� where the specification width is the difference between the

upper specification limit (USL) and the lower specification limit

(LSL) of the process.

� The process width is computed as 6 standard deviations (6σ)

of the process being monitored.

� A six sigma process is one in which 99.99966% of the products

manufactured are statistically expected to be free of defects

(3.4 defective parts/million) .

Page 65: Module 5 Spc

� Cp = 1 Process variability just meets specifications.

� Cp <= 1 Process is not capable of meeting

specifications .

� Cp >= 1 Process is capable of meeting specifications

Page 66: Module 5 Spc
Page 67: Module 5 Spc
Page 68: Module 5 Spc
Page 69: Module 5 Spc

Yes: No:

No:

No:

Yes:

Yes:

potentially capable

if re-centered

potentially capable

if re-centered

too wide

Page 70: Module 5 Spc

Problem

8/11/201570

� In a capabilty study of a lathe used in turning a shaft to a

diameter of 23.75 +-0.1 mm a sample of 6 consecutive

pieces was taken each day for 8 days . The values of ∑xbar=

190.156

� ∑R =0.54 .Construct the control chart and fnd out the

process capability of the machine

Page 71: Module 5 Spc

8/11/201571

d2= 2.534

A2=0.48

D3=0

D4=2

2

6Prd

RXocesswidth =

Page 72: Module 5 Spc

8/11/201572

� ucl = 23.802

� Lcl= 23.7322

� Cp=1.254

Page 73: Module 5 Spc

8/11/201573

� Problem

� The length of time customers of Statistical Software, Inc.

waited from the time their call was answered until a technical

representative answered their question or solved their

problem is recorded in Table 20-1.

� Develop a control chart.

� Does it appear that there is any time when there is too much

variation in the operation?

Page 74: Module 5 Spc

8/11/201574

Page 75: Module 5 Spc

8/11/201575

Page 76: Module 5 Spc

8/11/201576

Page 77: Module 5 Spc

8/11/201577

Page 78: Module 5 Spc

8/11/201578

Page 79: Module 5 Spc

8/11/201579

Page 80: Module 5 Spc

ATTRIBUTE CONTROL CHARTS

Page 81: Module 5 Spc

Attribute charts

� Many quality characteristics cannot be convenientlyrepresented numerically.

� In such cases, each item inspected is classified aseither conforming or nonconforming to thespecifications on that quality characteristic.

� Quality characteristics of this type are calledattributes.

� Examples are nonfunctional semiconductor chips,warped connecting rods, etc,.

Page 82: Module 5 Spc

Control Charts for Attributes Data

� p charts: proportion of units nonconforming

� np charts: number of units nonconforming

� c charts: count of nonconformities.

� u charts: count of nonconformities per unit.

Page 83: Module 5 Spc

Type of Attribute Charts

p charts

� This chart shows the fraction of nonconforming ordefective product produced by a manufacturingprocess.

� It is also called the control chart for fractionnonconforming.

np charts

� This chart shows the number of nonconforming.Almost the same as the p chart.

Page 84: Module 5 Spc

c charts

� This shows the number of defects or nonconformitiesproduced by a manufacturing process.

u charts

� This chart shows the nonconformities per unit produced bya manufacturing process.

Page 85: Module 5 Spc

p charts• In this chart, we plot the percent of defectives (per

batch, per day, per machine, etc.).

• However, the control limits in this chart are not based on

the distribution of rate events but rather on the

binomial distribution (of proportions).

Page 86: Module 5 Spc

Formula� Fraction nonconforming:

p = No of defects/n

� where p = proportion or fraction non conformities in the

sample or subgroup,

� n = number in the sample or subgroup,

Page 87: Module 5 Spc

P chart

n

pppUCL

)1(3

−+=

n

pppLCL

)1(3

−−=

inspected no total

defectives of No=p

Page 88: Module 5 Spc

8/11/201588

Sample

Nof pieces

inspected

No of defects

identified

1 300 25

2 300 30

3 300 35

4 300 40

5 300 45

6 300 35

7 300 40

8 300 30

9 300 20

10 300 50

Page 89: Module 5 Spc

8/11/201589

Sample no Sample sizeNo of defective

pieces

1 90 9

2 65 7

3 85 3

4 70 2

5 80 9

6 80 5

7 70 3

8 95 9

9 90 6

10 75 7

Page 90: Module 5 Spc

Example

Page 91: Module 5 Spc

P chart

Page 92: Module 5 Spc

np Chart

� When the subgroup size is constant, the chart constructed

for the actual no. of defectives rather than the fraction

defectives is called np-chart.

� Advantages

� np-chart is easier for operating personnel to understand.

� Inspection results are posted directly to chart without any

calculations.

Page 93: Module 5 Spc

8/11/201593

� A manufacturer uses a injection moulding to produce a plastic

insulation barrier. He inspects 100 barriers daily picked

randomly from the production and determines the no. of

defects by visual inspection. He wishes to use the data

accumulated during a 10 day period to construct an attribute

chart. The results of inspection are shown below.

(a) Plot np-chart and offer your comments

(b) What control limits would you recommend for the future

period.

Page 94: Module 5 Spc

8/11/201594

Page 95: Module 5 Spc

8/11/201595

Page 96: Module 5 Spc

8/11/201596

Page 97: Module 5 Spc

8/11/201597

Page 98: Module 5 Spc

c Chart

ccUCL 3+=ccLCL 3−=

•The procedures for c chart are the same as those for the p

chart.

•If count of nonconformities, co, is unknown, it must be found bycollecting data, calculating UCL & LCL.

samples of no Total

samples allin defects of No=C

Page 99: Module 5 Spc

8/11/201599

Applications of C and U chart

� Number of surface defects in a galvanized sheet.

� Number of imperfections in a certain area of cloth.

� Number of defective units in an air craft unit.

� Number of mistakes per unit.

Page 100: Module 5 Spc

8/11/2015100

� The following data refers to the no. of missing rivets on an

aircraft body noticed during preventive maintenance schedule.

(a) Compute the control limits for a suitable control chart.

(b) Plot the data and offer your comments.

(c) What value of C would you recommend for the future period

Page 101: Module 5 Spc

8/11/2015101

Page 102: Module 5 Spc

8/11/2015102

Page 103: Module 5 Spc

8/11/2015103

Page 104: Module 5 Spc

u Chart

� The u chart is mathematically equivalent to the c chart.

n

cu = ∑

∑=

n

cu

n

uuUCL 3+=

n

uuLCL 3−=

Page 105: Module 5 Spc

8/11/2015105

Listed below are the cloths produced on a daily basis in a small

textile mill and the corresponding number of imperfections

found in their bales is as follows.

(a) Use the data to estimate U .

(b) Determine the control limits and plot the data.

(c) What value of U1 would you recommend for the future period

Page 106: Module 5 Spc

8/11/2015106

Page 107: Module 5 Spc

8/11/2015107

Page 108: Module 5 Spc

8/11/2015108

Page 109: Module 5 Spc

8/11/2015109

Page 110: Module 5 Spc

8/11/2015110

� Determine the control limits for the following data using

suitable control chart. Plot the data and offer your

comments.

Page 111: Module 5 Spc

8/11/2015111

Page 112: Module 5 Spc

8/11/2015112

Page 113: Module 5 Spc

8/11/2015113

Page 114: Module 5 Spc

8/11/2015114

Castings

No of Defects

1 2

2 4

3 1

4 5

5 5

6 6

7 3

8 4

9 0

10 7

Page 115: Module 5 Spc

Nonconformity Classification

� Critical nonconformities

� Indicate hazardous or unsafe conditions.

� Major nonconformities

� Failure

� Minor nonconformities

Page 116: Module 5 Spc

Advantages of attribute control charts

� Allowing for quick summaries, that is, the engineer may simply

classify products as acceptable or unacceptable, based on

various quality criteria.

� Thus, attribute charts sometimes bypass the need for

expensive, precise devices and time-consuming measurement

procedures.

� More easily understood by managers unfamiliar with quality

control procedures.


Recommended