Scribd
Upload
73 views

Lecture4 Control Charts

This document discusses variable control charts, including X-bar and R charts as well as X-bar and s charts. It provides examples of how to construct control charts using sample data and determine control limits. Key steps include calculating the central line, determining trial control limits using factors like A2 and D4, and interpreting the charts to identify whether the process is in or out of control. Revising the charts involves eliminating out of control points and recalculating limits. The goal is to distinguish random variation from non-random patterns that could indicate a problem needing correction.

Uploaded by

Melissa John
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
73 views

Lecture4 Control Charts

This document discusses variable control charts, including X-bar and R charts as well as X-bar and s charts. It provides examples of how to construct control charts using sample data and determine control limits. Key steps include calculating the central line, determining trial control limits using factors like A2 and D4, and interpreting the charts to identify whether the process is in or out of control. Revising the charts involves eliminating out of control points and recalculating limits. The goal is to distinguish random variation from non-random patterns that could indicate a problem needing correction.

Uploaded by

Melissa John
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

STATISTICAL QUALITY CONTROL AND IMPROVEMENT

Lecture 4 Control Charts Shashikant Sathaye Polytechnic Institute of NYU

Variable Control Charts


How to Use the Process Control Charts
A process is said to be in control when the performance of the process falls within the statistically calculated control limits and exhibits only random variation due to chance causes Process under control does not mean that the process meets design specifications Identifying Patterns:
Trends are changes in levels; for e.g the center line might be drifting downwards Change or jump in level can be detected if a consecutive set of observations are consistently above(below) the center line.

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 2

Variable Control Charts


Example: Printer Assembly
An assembly area has been experiencing serious delays in the manufacture of computer printers. As a quality assurance manager you are asked to investigate and determine the cause of delays and fix the problems. Given the limited time, you get together with the representatives from all the departments and develop an cause and effect diagram Set the quality characteristic to measured length of a shaft

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 3

Variable Control Charts


Example: Printer Assembly
Choose a rational sub-group to sample needs to be homogeneous
Sub-group should be at least 4. Normally use range R for estimating std dev. If sub-group size is more than 10, then use s.

Collect the data Determine the trial central line for the X chart. X= Xi m

where X = average of the subgroup averages Xi = average of the ith group m = number of subgroups
Statistical Quality Control and Improvement Polytechnic Institute of NYU Lecture 4 MN611 4

Variable Control Charts


Determine the Trial Control Limits for the Xbar Chart UCLx = X + 3 x LCLx = X - 3 x Since the std deviation, is not known and we have a small sample size for each subgroup (4), we use the range R as a proxy for the std deviation and a factor A2 as the multiplier for the 99% confidence limit. A2 depends on the sample size. Thus UCLx = X + A2R LCLx = X A2R So for our example: X = (11.99+12.00,+12.00)/21 = 11.99 and R = (.08+.07_.04+.+.04)/21 = 0.05 UCLx = X + A2R = 11.99 + .577(.05) = 12.02 LCLx = X A2R = 11.99 - .577(.05) = 11.96

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 5

Variable Control Charts


Determine the Trial Control Limits for the R Chart R = Ri = .05 m The upper and lower control limits are determined using another set of multiplier D3 and D4; for a sample of n=5, D3 and D4 are 0 and 2.114. Thus UCLR = D4R = 2.114(.05) = 0.11 LCLR = D3R = 0(.05) = 0

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 6

Variable Control Charts


X
bar
12.04 12.02 12.00 11.98

UCLr CLine LCLr

C H A R T R C H A R T

11.96 11.94 11.92 11.90 11.88

0.12 0.10 0.08 0.06 0.04 0.02

UCLr

CL

LCLr

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 7

Variable Control Charts

Examine the process: Control Chart Interpretation


Avoid misinterpretations
Blaming people for problems they cannot control Spending time and money on process adjustments that are not necessary Asking for workers to improve when it the equipment that needs fixing

State of Process Control Process is under control when the performance falls within the statistically calculated control and exhibits only chance or random causes. +/-3 limits 99.73% of the measurements will be within the limits. Based upon the normal curve

Region A = 68.3% Region B = 27.2% Region C = 4.2%

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 8

Variable Control Charts

Examine the process: Control Chart Interpretation Thus the basic guidelines for a process under control are:
Two thirds of the points are close to center A few points on or near the center line Points equally spread around the center No points outside the control limits No patterns or trends on the chart

The range chart reveals the amount of variation in the process; so look for amount of variation in addition to the rules listed above. Types of patterns (non random variations)
Trends (upward or downward) Change or shift in level Runs Recurring cycles Non homogenous data

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 9

Variable Control Charts

Revising the charts


R chart should be under control (it appear so) Isolate the cause of out of control and fix the problem
12.04

X
ba r

12.02 12.00 11.98 11.96

UCLr CLine LCLr

C H A R T R C H A R T

11.94 11.92 11.90 11.88

0.12 0.10 0.08 0.06 0.04 0.02

UCLr

CL

LCLr

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 10

Variable Control Charts

Revise the charts


Eliminate the out of control points from both the charts and re-calculate the center line and control limits Eliminate the three points Xnew = and Rnew =

Xi - Xd m - md Ri - Rd m - md

= (251.77-11.94-11.95-11.95)/(21-3) = 12 = Xo

= (1.06-.04-.05-.6)(21-3) = .05

New Control Limits o = Rnew/d2 = .05/2.326 = 0.02 UCLx = Xo + A o = 12 + 1.342(.02) = 12.03 LCLx = Xo - A o = 12 - 1.342(.02) = 11.97

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 11

Variable Control Charts


Revising the control charts
12.04 12.02 12.00 11.98 11.96 11.94 11.92

UCLx UCLo Center Lineo LCLo

11.90 11.88

LCLx

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN6113 - 12

Variable Control Charts


Xandscharts
schartismoreprecisethanRchartsbutrequirelargersample subgroupsize(>10) ThecontrollimitsarecalculatedinasimilarmannerasXandR chartsexceptthatthe controllimitsarebasedupons. s=

si m

=(.03+.029+..+.011)/21=0.414/21=.02

UCLx =X+A3s=11.99+1.427(.02)=12.02 LCLx =XA3s=11.99 1.427(.02)=11.96 UCLs =B4s=2.089(.02)=.04 LCLs =B3s=0(.02)=0

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN6113 - 13

Variable Control Charts (Xbar and s charts)


X
bar
12.04 12.02 12.00 11.98

UCLr CLine LCLr

C H A R T S C H A R T

11.96 11.94 11.92 11.90 11.88

0.05 0.04 0.04 0.03 0.03 0.02 0.02 0.01 0.01

UCLs

CL

LCLs

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 14

Variable Control Charts

Statistical Quality Control and Improvement

Polytechnic Institute of NYU

Lecture 4 MN611 15

You might also like