Notes -POM- Module 7

Production and Operations ManagementModule 7: Quality Management

Definitions

Quality

Degree to which a set of characteristics of a product or service meet requirementsCharacteristics may be subjective or objectiveSubjective characteristics may be poor, good, excellent etc.Objective characteristics may be defective parts per million (PPM), complaints, Cost of poor Quality)

Characteristics may be:– Physical ( e.g.. Mechanical, electrical, chemical)– Sensory ( e.g.. Smell, touch, taste, sight, hearing)– Behavioral (e.g.. courtesy, honesty)– Temporal (e.g.. Punctuality, availability, reliability)– Ergonomic ( related to human safety)– Functional ( performance oriented)– Others (e.g.. maintainability, reliability, spares support)

Inspection:Inspection deals with verification of products to ensure that product produced meet specified requirements. Main purpose of Inspection is to segregate good products from bad.Inspection may at receiving stage, in process stages and final stageInspection may be inspection with instruments , testing , visual Inspection may be 100 %, sampling or Audit . extent of inspection depends on cost and risk of bad products being accepted.

Quality ControlQuality control covers activities that are performed to ensure that product meet requirements. These include right selection of materials, Statistical process control, calibration of instruments, verification of tooling, setup inspections. Quality control activities cover inspection also.

Statistical Quality ControlStatistical Quality control deals with the application of statistical techniques for control of processes to produce quality products and also for inspection

Variability of a process (Process variation)Each work station in a manufacturing cycle converts inputs to outputs. These processes convert material inputs to value added outputs. The characteristics of Products produced may be:

Dimensional characteristics (eg Length, thickness, diameter) Form characteristics ( eg. parallelism, ovality) Physical and chemical properties ( eg. Hardness, chemical composition)

If a batch of products is produced, none individual items are not identical. There will always be very small differences. Some times, these differences are apparent only when measurements are taken with high accuracy instruments.

Variation in the batch= maximum dimension – minimum dimension

These variations are due to two broad causes Chance causes or Common causes Special causes or Assignable causes

Examples of chance causes are: slight variation in materials equipment variations (eg. Run out, play, backlash ,geometric inaccuracies

positional accuracy, repeatability) clearances in fixtures minute tool wear variations in environment (temperature, humidity) slight variation in measurements

Characteristics of chance causes are: many individual causes are present at the same time any one cause results in only a very a small amount of variation cannot be economically eliminated Output of process follows predictable pattern ( bell curve/ Normal curve)

Examples of Special causes are: Tool not sharpened when worn out batch of defective material component not located properly power failure excessive variation in measurements machine malfunction

setting disturbedCharacteristics of Special causes are:

Consists of one or few individual causes any cause can result in large variation easy to detect and generally economical to eliminate no specific pattern n the Output

Variation in the process is minimum when only chance causes are present. Also the distribution ( frequency diagram)out put exhibits normal curve.

Distribution pattern (only common causes are present)

variation in a process is the total variation that can be expected from the process when very large quantities are produced.

Variation in the process is be computed based on the output values of a batch. Normally the batch should be more 100 components .i.e. there should be at least 100 readings to calculate the process variation.

Histogram or frequency diagram is drawn to make sure that only chance causes are present. In other wards the process is under control (statistical control)Suppose N readins are taken, then

Process Variation

Average

Standard deviation ( sigma) ( σ ) is calculated from the formula

σ = root of ( (( M-X1)2 + ( M-X2)2 + ( M-X3)2 …….( M-XN)2 )/ N-1))

Where M= average of N readings X1, X2 ,X3 ……. XN are individual readings

Process variation = 6 times σ or 6 σ

Property of Normal curve

In manufacturing operations less variation means better quality.If special causes are present , the variation will be more than 6 σ. Or the pattern would change. This property is used in construction and use of control chats for statistical process control.

The out put of a process has two properties. Process average and process variation (6 sigma)Out put of a process should meet process requirements. Requirements are given in specifications or Drawings. A typical specification has a mean and Tolerance.Example: Length to be 100.0 +0.6mm means product is acceptable if the length is 100.0 or 100.6mm.

Average

Process variation (6 σ)

68.3% of readings would lie in this range

99.7% of readings would lie in this range

95.4% of readings would lie in this range

3 out of 1000 reading may fall outside +/- 3 σ limits

Alternatively specification can be given as 100.0/100.6. This gives the lower and upper specification limits.

Difference between upper specification and lower specification is the tolerance.

Two process measurements are normally made to indicate the capability of the process to meet requirements. Namely:

Process capability Cp Process capability index Cpk

Example:

Process capability = Tolerance/ process variation For the above example, the process variation is 0.3:Process capability = 0.6/0.3 = 2.0

Process capability IndexThis takes into account both tolerance and specification mean.Cpk is calculated for both upper and lower specification limits and lower of the two, is the Cpk

Cpk ( Upper)= (upper specification- process average)/ 3 sigma = (100.6-100.4)/.15 = 1.33 Cpk ( lower)= (process average –lower specification)/ 3 sigma = (100.4-100.0)/.15 = 2.66

100.0LSL 100.6

USL

100.3Specn. mean

100.4Process average

Process variation(6 σ)=0.3

Tolerance =0.6

Cpk of the process is 1.33.

If the process capability index is equal to 1.33 , the process is considered acceptable. More than 1.33 is desirable. This can take up variations in process average without producing rejects on the long run.

Control chartsControl charts were developed by Walter Shewhart in 1920. The chart is based on the principle that the variability of a process is minimum when the process is running under chance causes (common causes) only. Trial batches are produced and process variation ( 6 sigma)is established . Control charts are prepared with control limits calculated on sample sizes . Sample sizes could be 2 to up to 9 . At fixed intervals ( eg, 30 min, one hr, 4 hrs based on production quantity) samples are taken and sample averages are plotted on the chart.

When the process is under control, all the points will be with in the limits No unnatural patterns will be present Distribution of points will be denser in the centre and very infrequent at the limits.

Any change in the above indicates presence assignable/ special causes (out of statistical control). The process is stopped and actions are taken to eliminate the causes of variation.

Mean

Time

LCL

UCL

Construction of X bar –R chart

Step 1 - Choose quality characteristic to be controlled. Step2 - determine sample size ( 3 to 5 is ideal) and no. of samples.Step 3 - Collect samples at fixed intervals and note readings ; no. of samples depends on production rate and total duration of manufacture.Step4 -select No. of subgroups ( K ) Normally total no. of readings required is 75 to125Step 5 – Calculate mean and range for each sample

Example : sample size =3 : No. of subgroups=26 parameter for charting -bond strength in grams

Sub group no.

Sample 1 Sample 2 Sample 3 Sub group Mean(X-bar)

Range ( R )

1 6.5 6.5 5.0 6.0 1.52 6.5 6.0 7.0 6.5 1.03 7.0 5.5 5.5 6.0 1.54 4.0 3.0 3.5 3.5 1.05 2.0 0.0 1.0 1.0 2.06 6.0 6.0 6.0 6.0 07 6.0 4.0 4.0 4.6 2.08 6.0 4.0 6.0 5.3 2.09 5.5 5.0 4.5 5.0 1.010 7.0 6.0 6.0 6.3 1.011 4.0 3.0 3.5 3.5 1.012 4.0 3.5 3.0 3.5 1.013 5.0 3.5 5.0 4.5 1.514 6.0 4.5 4.5 5.0 1.515 4.0 6.5 5.5 5.3 2.516 4.0 7.0 7.0 6.0 3.017 5.0 6.0 4.0 5.0 2.018 3.0 6.0 6.0 5.0 3.019 5.5 5.0 7.5 6.0 2,520 7.0 6.0 5.0 6.0 2.021 5.0 5.0 5.0 5.0 022 5.0 2.5 3.5 3.6 2.523 4.0 5.5 5.0 4.8 1.524 4.0 0 2.0 2.0 4.025 5.0 3.0 3.5 3.8 2.026 2.5 5.0 3.0 3.5 2.5

Step 5 – Calculate mean of subgroups ( X-double bar)= (6.0 + 6.5 + ………..+ 3.8 + 3.5 ) / 26 = 4.72

Step 6- Calculate mean of ranges ( R-bar)=( 1.5 + 1.0 + ………… + 2.0 + 2,5) / 26 = 1.75

Step 7 - Calculate control limits for X-bar chart and R –chart Use statistical tables below:

Sample size

A2 D3 D4

2 1.88 0 3.273 1.02 0 2.574 0.73 0 2.285 0.58 0 2.11

For X-bar chart: ULC = X-double bar + A2 x R-bar = 4.72 + 1.02 x 1.75 = 6.50LCL = X-double bar - A2 x R-bar = 4.72 - 1.02 x 1.75 = 2.94

For R chart: ULC = D4 x R-bar = 2.57 x 1.75 = 4.50LCL = D3 x R-bar = 0 x 1.75 = 0

Step 8 – test for HomogeneityCheck sample averages or Ranges fall outside the control limits for any subgroups. sub group averages for no. 5 and 24 fall below lower control limit all values in Range are within control limits

Step 9 – remove subgroups which are outside control limits .

Step 10- Recalculate control limits

Modified X- double bar = 4.99 For X chart : Modified UCL = 6.67 : modified LCL= 4.21For range chart ; Modified ULC = 3.31 : modified LCL=0

Step 11- check for Homogeneity. Repeat the process till all reading are homogeneous.In the example , All 24 subgroups are with in control limits.

Step 12- construct control chart for X-bar and R chart

Types of control charts are: 1. For variable measurements (Quantitative) ( eg. Weight, dimensions):

X-R chart ( mean -Range chart) - constant sample size – most popular X-MR chart ( mean – moving range chart)- for single value per batch Median- range ( middle value –range chart) – for easy calculations X-S chart ( mean –standard deviation chart – for large sample sizes

2 For Attribute measurements (Qualitative) (OK/not OK, Good/bad): np chart, - no of defectives – for constant sample size p chart, - percentage defectives – for varying sample sizes c chart – no of defects - for constant sample size u chart – no. of defects per part - for varying sample sizes

Control chart for attributes

X-bar chart

UCL = 6.67

Mean =4.99

LCL = 3.31

R chart

UCL = 4.21

Mean =1.64

LCL = 0

Sample size = n : No. pf subgroups = K : Number of defectives in sub group = cp= fraction defective = c/n ; p is calculated for each sub group ( p1=c1/n , p2=c2/n ……)p-bar ( average/ mean ) = (c1 + c2 + …….) nK

UCL = n x p-bar + 3 Sqrt ( n x p-bar ( 1 - p-bar))LCL = n x p-bar - 3 Sqrt ( n x p-bar ( 1 - p-bar))

Acceptance sampling

100 % Inspection is costly and time consuming when quantity to be inspected is large. Sampling inspection is the best way of estimating the quality of incoming or outgoing lots.

Acceptance sampling involves taking a random sample from the lot and by inspection of the sample and after inspection to determine whether to accept the lot or not.

Sampling may be: Single sampling Double sampling Sequential sampling

The sampling plans are based on average outgoing quality levels required on the long run. Based on the these quality levels sample sizes are determined and given in statistical tables. IS 2500 is a standard which gives such a table.

np chart

UCL

Mean = p-bar

LCL

Depending on the average outgoing quality levels the tables gives values for lot size, sample size, quantity of defectives that is acceptable.

There is always a risk of rejecting good lots and risk of accepting bad lots in any sampling plan.

The probability o accepting bad lots is called consumer’s risk . bad lot are defined by LTPD value ( lot tolerance percent defective) and The probability o rejecting good lots is called producer’s risk. Good lots are defined by Acceptable quality levels ( AQL)

The graph which gives the characteristics of the sampling plan indicating producers risk and consumers risk is known as OC curve

Average Outgoing Quality level (AOQL) is the percent defectives after lots are accepted and replacement of all defective items on rejected lots.

Single samplingSample is taken only once. If the no. of defects are more than acceptable then the lot is rejected and 100 % inspection is carried out.Double sampling

OC curve

Actual percentage defective

Probabality of acceptance

AQL=0.1 LTPD=0.5

0.10

0.95

Depending on the number of defectives in the first sample , second sample is drawn and decision is taken. Therte will be an acceptance number c1 and a rejection number c2 fixed during the first sampling. If the value is between this , second sample is taken Sequential sampling Allows more samples to be drawn based on number of defectives in the previous sample .until the lot is accepted or rejected

Quality Circles

Quality circle (QC) is a small group of people who carry out quality improvement activities within their work area. The group may consist of 3 to 5 person normally. Features of Quality circles are:

1. QC activities performed by operating level employees.2. Voluntary participation3. Small group of persons in the same work area or doing similar type of work4. Identify, define and solve problems related to work area5. QC members meet regularly( every week for an hour in normal)6. Leads to improved performance I the work area7. motivates and enriches work life8. Nature of the problems are of such that they can be solved by themselves with

little help form management9. Good work is recognized by management

Quality circles originated in Japan and has spread to other countries. Some of the organisations in India which have very active QC circles are BHEL, BEL, Canara Bank, Indian airlines, Apollo Hospital. Quality Circle Forum of India ( QCFI), a Non-profit national body , Promotes Quality circle movement in India. It has more than 2500 members and more than 25 chapters in various locations in India. QCFI organises Quality circle conventions. QC presentations by circle members and awards are a part of the convention.

Some of the don’ts in QC circles: QC members should not be chosen by management Problems should not be given by management Management persons should not be members.

Problems taken up by QC s are not restricted to Quality. Productivity improvements, cost reduction, safety, Housekeeping etc. could also be taken up. QCs are not forum for airing grievance or demands. Quality circles are not replacement for task forces, suggestion schemes,

Key benefits of Quality circles are:

1. Improved quality, productivity, cost reduction, House keeping, safety etc. which are the objectives for improvement

2. Improvement of Human relations in work area3. Effective team work4. Improved communication between employees and management5. Problem solving capabilities of members6. Leadership development among members

Implementation of QCsTop management must inform their decision to all employees about the implementation of QC programSteering committee to be formed with members from various departments Steering committee should define:

Goals of QC program Long rage plan for programs Select facilitator and commit resources Encourage and provide resources for QC meetings Participate in QC presentations

Facilitator: Facilitator forms the link between top management and QCs. Some of the duties of facilitator are:

Co-ordination of training activities for members. To get support from top management and steering committee To provide resources.

Circle leader: Circle leader is the leader of the circle and must: Inform status of activities to management conduct meetings regularly Monitor circle activities with regard to plan Maintain enthusiasm and motivate members

Circle members: Circle members must participate actively in meetings and discussions. They should develop a good attitude towards Quality, productivity and improvements.

Five major steps in implementation of QC program are Top management orientation Training for middle management Training for facilitator, leaders and members Pilot testing in selected areas Company –wide implementation

Subjects for training for Quality circle include: Back ground and concept of Quality circles Structure of QC Problem solving methodology

7 QC tools Mock circle meetings Presentation skills

Total Quality management

TQM is a Management philosophy and company practices that aim to harness the human and material resources in most effective way to achieve objectives of the organization

Total in “Total Quality management” means All Interested parties are considered All requirements are addressed All activities of the organization are covered All employees are involved

TQM-Fundamental concepts

CommitmentCommitment to TQM by top management is essential. Promotion of this concept at all levels and all activities of the organization is fundamental for success. Every one should be aware of his/her customers (internal and external)

Customer satisfactionUnderstanding customer needs and expectations and striving to meet them should be the key objective

Quality lossesQuality looses are losses caused by the failure to utilize most effectively the potential of human, financial and material resources in a process.

loss of customer satisfaction Loss of opportunity to add more value to customers,

organization and society Loss due to waste or misuse of resources

Participation by allAbilities of all members in the organization should be fully and effectively utilized.

Process measurementsProcess measurements to be applied to all organization activities

Continuous improvementsThe means of improvement to people and processes performance need to be Continually sought and monitored.

Problem identificationProvision for identification and resolution of potential and existing problem son a continuing basis is essentialAlignment of corporate objectives and individual attitudeRemoval of prejudices and restrictive approvals that inhibit effectiveness of the organization needs continuous attention

Personal accountabilityRecognition of individual responsibility and authority should be accepted by all

Personal developmentThere should be continuous appraisal, training and development of individuals at all levels

TQM- implementation 1. Policy and strategy of the organization

MissionEstablish mission statement, corporate objectives, strategy for achieving these objectives and a business plan.Leadership and commitmentHave visible and sustained commitment starting from chief executive and extending to every member.Divisional objectivesEstablish objectives for each level of organization with roles and responsibilities

2 Management of the organization

Organization structureEstablish effective organization structureManagement systemEstablish, audit and keep under review an effective management systemInformation systemEstablish a planned information system through out the organization

3 Improvement of the organizationWorking environmentStructure Physical environment and relationships between individuals and the organization1Measurement of performance

Establish measures of performance of individuals or teams involved in each process

Improvement objectivesImprovement goals to be closely integrated with corporate objectivesImprovement plansEstablish plans for improvement of products, service or process quality, safety, environment impact, dependability and customer satisfaction at all levelsMonitor and reviewEnsure that all plans, targets, and measure compliment each other.Review results of improvement plans to measure effectiveness

Some recent trends in TQM

Improve customer satisfactionImplement auditable Quality management system

Implement ISO9001, QS9000, TS16949, AS9000 as applicable

Improve environmental performance Implement ISO 14001

Improve safety and Hygiene Implement OSHAS 18001/ ISO 15001

Enhance social responsibility Implement SA 8000

Identify Improvement opportunities Conduct value stream mapping

Extensive use of tools and techniques in problem solving

7 QC tools7 Management tools

Improve supplier performance Supplier QMS development

Reduce lead time Flexible Manufacturing systemsComputerizationUse of information technology

Reduce inventory, reduce lead times JIT systemIntegration of suppliers

Effect Quantum leap/ break through improvements

Bench marking, Six sigma initiativesBusiness process re-engineeringTotal productive maintenance(TPM)Lean manufacturing

Participation by all employees Suggestion schemes5S House keepingQuality circlesQuality teams

Improve product quality Quality function deploymentDesign of Experiments

Improve planning process Policy deployment ( Goals, objective, Tasks)

Barriers in implementation of TQMBarriers to implementation of TQM may be divided into two groups. Both of them can be tackled by systematic education, training and participation.

Organizational barriersThese consist of lack of senior management commitment, unwilling ness and non-involvement of middle level management. Staff treating themselves as experts and not amenable to suggestions from line people..

Behavioral barriersThese may be also called as attitudinal barriers. Artificial organizational structures, negative attitude to changes and improvements, dislike because of previous experiences arte some of the factors.

Effects of TQM implementation in a company

# Aspect Before After1 Top management

commitmentNot visible Highly visible

2 Policy clarity No policy Transparent , detailed3 Participation in

improvementsVery few persons All levels

4 Involvement of people < 10% >80%5 Communication No or poor

communicationEffective communication at all levels

6 Recognition Seldom Always7 Customer satisfaction low high8 Cost of poor quality high Very low9 Training activities Almost nil Continuous, every one10 Performance

measurementsNot focused Focused and targeted

11 Business results Stagnant or declining Steady improvement12 Suppliers Outsiders keep them in

dark, squeeze them.Willing to work as extension of the company