Unit 9 - SPC

Question 1

Explain the history of SPC

Answer 1

First proposed on 1931, used by US industry during WW2 then mostly forgotten, taught to Japanese industry in 1950’s by Dunning, rediscovered to the west in 1980’s

Question 2

Explain process variation

Answer 2

All manufacturing processes exhibit variation
Sampled data on a product reveals variations in attributed about some value
This is called spread/variability
Common causes: variation inherent in a process, can only be eliminated through improvements in the system (random error)
Special cases: variation due to identifiable errors, can be modified through operator or management action (consistent error)

Question 3

Discuss how control charts are used in SPC

Answer 3

Used to monitor processes
Random variation due to common causes are normal for a stable process
Variations due to special causes are ide tidied and can be corrected

Question 4

Discuss sampling and control limits

Answer 4

Process outputs are sampled
Control limits are determined to specify the normal variation of the process
If measure samples lie outside of the control limits this indicates a special cause
The control limits are 3 sigma above or below the process average

Question 5

Explain SPC

Answer 5

Take periodic samples from process
Plot sample points on control chart
Determine if the process is within limits
Identify quality problems so that corrective action can be taken

Question 6

Is us the difference between process variation and sample average variation

Answer 6

If we calculate the average and standard deviation for the overall process over three trials, we can obtain spastically distribution for both the process and sample averages

Question 7

Describe the two types of data

Answer 7

Attribute: product characteristic evaluated with a discrete choice (yes/no)
Variable: product characteristic that can be measured (length, velocity)

Question 8

What is an X-bars chart

Answer 8

A control chart that uses the average or mean of a sample

Question 9

What is a range chart (R-chart)

Answer 9

A control chart that uses the amount of dispersion in a sample

Question 10

Plain how control charts are developed on the shop floor

Answer 10

They are based on in-control data
To simplify calculations, control limits are calculated from sample size, range data, and tabulated constants
Is non random causes are present, discard data

Question 11

Explain the two types of attribute control charts

Answer 11

P-chart: monitors fraction of defective parts
U-chart: monitors number of defects in a unit

Question 12

What determines a controlled or stable process

Answer 12

No sample point outside of limits
Most points near the process average
Equal number of points above and below centre line
Points are randomly distributed
Process runs in consistent and predictable manner
If variation is uncontrolled the. Either the process average is changing or the process average is changing or both

Question 13

What are the control chart rules

Answer 13

1. Any point above or below the control limits indicates a special cause
2. 7 points in a row above or below the centreline indicate a shift in the process mean
3. 7 points in a row trending up or down indicates a trend in the process
4. 2 out of 3 consecutive points near the control limits indicate a potential special cause
5. 4 out of 5 consecutive poins beyond 1 sigma from the centreline indicate a potential special cause
6. A run of 14 points alternating up and down indicate a systemic variation or oscillation
7. Isolated points far from the rest of data may indicate a one time anomaly
8. Any non random pattern such as cycles or stratification may indicate a special cause

Question 14

How are sample sizes for attribute and variable control charts

Answer 14

Attribute: 50 to 100 parts in a sample
Variable: 2 to 10 parts in a sample

Question 15

Explain process capability

Answer 15

The range of natural variability in a process is measured with control charts
The process cannot meet specifications if natural variation exceeds tolerances
Huge difference between 3 sigma and 6 sigma quality!

Question 16

Describe control versus specification limits

Answer 16

Specification limits: set by the product or process requirements
Product requirements are set by the customer
Process requirements are usually set internally

Question 17

When is a process considered capable

Answer 17

When the inherent variation of a process is less than the specification limits determined by the process caoeability index

Question 18

What is the difference between Cp and Cpk

Answer 18

Cp measures the potential capability of a process
Cpk measures the actual process capability, it estimates what the process is capable of producing considering that the process mean may not be centered between the specification limits

Question 19

How are Cpk values interpreted

Answer 19

A value less than zero indicates a process that is not producing between specified limits
A value of 1 indicates that the process meets customer requirements but with higher variability
Recommended values:
Existing process = 1.33 or higher
New process = 1.50 or higher
Six sigma process = 2.0 or higher

Question 20

What is the performance index

Answer 20

A measure of the overall performance of a process, a higher value of Pp indicates better process performance

Question 21

How are Ppk values interpreted

Answer 21

A value less than 0 indicated a process that is not producing between specified limits
A value of 1 indicated that the process meets requirements but with higher variability

Question 22

Compare Cpk and Ppk

Answer 22

Or ukase are the same, but Cpk is used for short term process analysis while Ppk is used for long term process analysis.
Cpk analyzes a subgroup of data, Ppk analyzes the entire dataset
Cpk is used to improve processes when the process is stable
Ppk is used to evaluate long term trends in performance of performance is an unstable process. Typically used for new process validation.

Question 23

What are the requirements of SPD

Answer 23

Involvement of workers
SPC alone is not always sufficient
SPC does not guarantee zero defects