Final - SPC Flashcards
quality inspection
focus on providing information
quality control
focus on monitoring and controlling -> SPC
quality assurance
management programs aimed at ensuring good product quality by setting minimum or desired levels of quality
TQM
total quality mgmt; mgmt philosophy about ensuring and improving product quality throughout the entire org
what are the 4 stages of quality management?
quality inspection, quality control, quality assurance, TQM
SPC (basic)
Statistical process control; stage 2 of quality management process; aims to: prevent poor product quality; Process improvement tool
when should quality inspection be done? (5)
upon receipt of resources
before transformation ops (especially bottleneck)
the first few items coming out of an automated operation
final inspection
customer complaints and returned goods
what are some factors in determining how much and how often quality inspection should be done?
complete inspection vs. sampling
cost of inspection vs. cost of not detecting defects
cost, time, and physical possibility
acceptance sampling
the method of randomly inspecting a sample of goods and deciding whether to accept the entire lot based on the results
what are the steps in acceptance sampling?
- take random sample from a lot (batch) of items
- test the sample items for the specified quality characteristics
- accept all items in the lot if
(# of defective items in the sample) < (maximum # of defective items allowed in a sample)
else reject all items in the lot
historic mentality
“some degree of poor quality will occur and that is acceptable” -> acceptance sampling is based on this standard and thus may not be the best method of quality control in today’s day and age, esp as bad quality grows increasingly untolerated
What is SPC (describe)?
Method of randomly inspecting a sample of goods and deciding whether the production process is in control
- Monitor the production process (data pattern)
- Provide a statistical signal when the process (quality) changes
What are the steps in SPC? IMPORTANT
- Define quality characteristic to measure
- Set up a control chart
- Take a random sample and plot the quality measure
- If sample point is not outside of control limits -> Step 5 is to check if any pattern exists. If it is, proceed to step 6.
- If pattern’s don’t exist, the process is in control. If it does, proceed to step 6
- Process is out of control; STOP the process until quality problem is identified and fixed
attributes
characteristics that are measured qualitatively, thus have DISCRETE values (COUNTABLE)
ex. defective/non-defective, # of scratches or blemishes
variables
characteristics that are measured quantitatively, thus have CONTINUOUS values (NON-COUNTABLE)
ex. weight, length, volume, temperature
What are the 4 different SPC charts that should be used when defining a quality characteristic to measure?
For attributes:
p-chart: Counting both defective and non-defective items
c-chart: Counting defective features in an item
For variables:
R-chart: Difference (range) between smallest and largest values aka max -> min
x-chart: Average
UCL
Upper control limit
LCL
Lower control limit
what is the norm for # of sigma limits?
3
random/normal variation occurs in
the range between the uCL and LCL (in control; average)
abnormal variation occurs in
observations falling outside the UCL or LCL range
what is the control limit standard?
3 sigma limits
what type of error is produced when control limits are too narrow?
type I error: random variation mistaken for an abnormal variation (belief quality is out of control when it’s actually in control) -> CONSERVATIVE
what type of error is produced when control limits are too wide?
type II error: abnormal variation may not be detected (belief quality is in control when it’s actually out of control) -> WORST CASE
sigma limits
limits on the variability of a process; used to identify when a process is operating outside of its normal range, which may indicate that there is a problem with the process that needs to be addressed
expressed as a number of standard deviations away from the mean
ex. if a process has a mean of 100 and a standard deviation of 10, the 3 sigma limits for that process would be 70 and 130. This means that if the process produces a result that is outside of this range, it is considered to be outside of the normal range and may require further investigation.
how are sigma limits calculated?
Based on standard deviation and mean, so (mean - (sd*#of sigma limits))
ex. if mean is 100 and sd is 10, with 3 being the # of limits, then 100 +/- (3*10) = 70 and 130
what does the p-chart measure?
the proportion of defective items
how to find out what the upper and lower bounds of a sample are?
take historical average (aka center line or p bar) and use UCL and LCL formulas
if hist. avg isn’t provided, use the sample average by adding up all the porportions s(p) divided by the # of samples
n = # of samples, z = # of standard deviations (sigma limits), typically 3
what does the c-chart measure?
NUMBER of defective features per item (not possible to count non-defective, therefore cannot compute proportion)
what is the sample size in a c-chart?
1 item per sample
what is the sample size in a p-chart?
typically 30-100 items per sample
If the LCL of a control chart is computed as a negative number, what should the actual LCL be?
- Ex. # of luggages cannot be negative, so the new LCL is 0 instead of the negative number computer
what does the R-chart measure?
dispersion on variability, as determined by subtracting the smallest value from the largest value in the sample
what is the sample size in an R-chart and x-chart?
2-10 items per sample
what does the x-chart measure?
sample average
What is A2, D3, D4, respectively?
A2 = mean value constant based on sample size; specifically developed to determine control limits for x-chart
D3, D4 = range value constants based on sample size; specifically developed to determine control limits for R-chart
Why is it important to use both x- and R-chart in SPC?
The x-chart is used to monitor the process mean, while the R-chart is used to monitor the process variability. Both of these factors can affect the quality and consistency of the product or service being produced.
For example, if the x-chart shows that the process mean has shifted significantly, this could indicate a problem with the process or a change in the input materials. On the other hand, if the R-chart shows an increase in the process variability, this could indicate a problem with the process consistency or a change in the process conditions.
By using both charts together, you can identify and address these issues in a timely manner, which can help to improve the overall performance and quality of the process.
What are the steps for using x- and R-chart together?
- Calculate x and R of each sample (average, and largest value-smallest value)
- Calculate UCL and LCL of each (formula sheet, reference the chart for A2, D3, D4 values based on sample size etc.)
- Plot points (calculated in step 1), draw limit lines, and identify if any points are out of bounds
what is the objective of pattern tests?
to detect a non-random pattern within the control limits (step 5)
what two parts comprise a pattern test?
run tests (patterns 1-3), zone test (patterns 4-5)
run
a sequence of observations with a certain characteristic
how many zones are there in a control chart?
3 for 3 sigma limits
Where are Zones A-B in a control chart?
Zone A is outer bounds of chart (at UCL and LCL)
Zone B is next middle tier
Zone C are the two sections closests to centre line
Visually:
A
B
C
C
B
A
How is the zone width determined?
(UCL-LCL)/3 (or the sigma limit) , but subtract center line from UCL if you’re dividing into 6 zones
process capabnility
the ability of a process to satisfy a product’s or service’s tolerances (CUSTOMER PERSPECTIVE)
tolerances
design specifications that reflect customer requirements
UTL and LTL
upper and lower tolerance limits; not statistically determined and not a result of production processes
the production process must have what 2 things to be of quality?
process control (check SPC), AND meeting design specifications (check process capability)
process capability index
a way to assess process capability by comparing specification width (UTL-LTL) against process width (UCL-LCL); denoted by ‘Cp’
The production process is capable of meeting design specifications if
Cp = 1 (spec width = process width)
The production process is capable of exceeding specifications if
Cp > 1 (spec width > process width)
The production process is not capable of meeting specifications if
Cp < 1 (spec width < process width)
