M2 STATISTICAL QUALITY CONTROL Flashcards
identified portion of a batch having uniform character and quality within specified limits
Lot/Batch
representative/portion
Sample
Look at all 5,000 shells
100% inspection
Don’t look at any, put the whole shipment into stock
0% inspection
Look at some of them, and if enough of those are good, keep the lot
Acceptance sampling
status
Lot disposition
decision
Lot sentencing
- Testing is destructive
- Cost of 100% inspection is high
- 100% inspection is not feasible
(require too much time) - If vendor has excellent quality history
8
Statistical Quality Control
Why Acceptance Sampling and Not 100% Inspection
- Less expensive
- Reduced damage
- Reduces the amount of inspection error
Advantages of Sampling
- Risk of
– accepting “bad” lots - Consumer’s Risk
– rejecting “good” lots - Producer’s Risk
- Less information generated
- Requires planning and documentation
Disadvantages of Sampling
accepting “bad” lots
Consumer’s Risk
rejecting “good” lots
Producer’s Risk
Indicated the number of units of product from each lot or batch which are to be inspected (sample size or series of sample sizes) and the criteria for determining the acceptability of the lot or batch (acceptance and rejection numbers)
Sampling Plan
The units selected for inspection should be chosen at “random.”
Random Sampling
unbiased
Random Sampling
When we took the capsule shell sample, we didn’t put them back into the lot during sampling, i.e., we didn’t replace them.
Sampling with/without Replacement
acceptance sampling
we sample without replacement
each piece in the lot has equal probability of being in the sample
simple random sample
he lot is divided into ‘H’ groups, called “strata.”
Each item in the lot is in one and only one stratum
stratified sample
Sampling plans are typically set up with reference to–
Acceptable Quality Level (AQL)
is the base line requirement for the quality of the producer’s product
Acceptable Quality Level (AQL)
– the poorest level of quality (percent nonconforming) that the process can tolerate
Acceptable Quality Level (AQL)
Scenario:
– N = 100 products
– AQL = 10% (max. 10 defectives allowed)
– If no. of defectives < 10 → Accept
– If no. of defectives > 10 → Reject
Acceptable Quality Level (AQL)
Scenario:
– N = 100 products
– AQL = 5% (max. 5 defectives allowed)
– If no. of defectives < 5 → Accept
– If no. of defectives > 5 → Reject
Acceptable Quality Level (AQL)
Which one has higher standard of quality in the production process?
Higher or lower AQL??
Lower
- Used to measure acceptable levels of quality of the products inspected
- Widely used to decide whether to accept a production lot without checking every single item.
MIL-STD-105E
– The most recently published version is MIL-STD-105E
- Notice 1 cancelled the standard and
refers DoD users to ANSI/ASQC Z1.4-1993
MIL-STD-105
AQL and defect classification Defects detected during inspections are generally classified in 3 categories
- Critical defect
- Major defect
- Minor defect
An undesirable characteristics of a product
Defects
is a unit of a product which contains one or more defects
DEFECTIVE
Classification of Defects
* According to Seriousness or Gravity
- Critical defect
- Major defect
- Minor defect
can endanger life or property, may render the product non-functional
Critical Defect
renders the product useless
Major Defect
slight variations
Minor Defect
Classification of Defects
- According to Measurability
–Variable Defect
–Attribute Defect
single, measurable quality characteristic
Variable Defect
- Not represented numerically
- Defective or non-defective
- Conforming or Non-conforming
Attribute Defect
Classification of Defects
- According to Nature
–Ocular Defect - visible
– Internal Defect – not seen but present
–Performance Defect - function
Variations
– Cannot be identified and unavoidable
– Common and inherent on the process
– Cause slight differences in process variables like diameter, weight, service time, temperature, etc.
Random
Variations
– Assignable
– Causes can be identified and eliminated
– Typical causes are poor employee training, worn tool, machine needing repair, etc.
Non-random
(Assessing the Quality of Results)
Statistics
Sources of Variations
– Variation between supplies of same substance
– Between batches from same supplier
– Within a batch
Materials
Sources of Variations
– Variation of equipment for the same process
– Differences adjustment of equipment
– Aging and improper care
Machines
Sources of Variations
– Inexact procedures
– Inadequate processes
– Negligence by chance
Methods
Sources of Variations
– Improper working conditions
– Inadequate training and understanding
– Dishonesty, fatigue and carelessness
Men
- It is impossible to perform analysis without any error or uncertainties.
- The TRUE value of a measurement is never known exactly.
- Replicates are samples of the same size that are analyzed in exactly the same way.
Statistics
The TRUE value of a measurement is
never known exactly
are samples of the same size that are analyzed in exactly the same way.
Replicates
It is impossible to perform analysis without any
error or uncertainties
(Average)
Mean
central or “best” value
Mean and median
(middle result when arrange in order)
– Odd numbered samples → the middle
– Even numbered samples → average of the middle pair
Median
– Closeness of measurements with each other
– Reproducibility of measurements
PRECISION
– STANDARD DEVIATION
(s or SD)
* The higher the s, the less precise the measurements are.
– VARIANCE (s2)
– COEFFICIENT OF VARIANCE (CV)
– RANGE (SPREAD)
PRECISION
– Closeness of measurement to its TRUE (accepted) value
ACCURACY
- Composed of:
– Central Line
– Upper Line for the UCL
– Lower Line for the LCL - Can conclude if variation is
– CONSISTENT (in-control)
– UNPREDICTABLE (out of control)
Control Charts
– ERROR
* ABSOLUTE ERROR
* RELATIVE ERROR
ACCURACY
- means of visualizing variations that occur in the central tendency and dispersion of a set of observations
- used to study how a process changes over time
CONTROL CHARTS
TYPES OF CONTROL CHARTS
- ATTRIBUTES
– p-chart
– c- chart - VARIABLES
– X-chart
– R-chart
– X-chart
– R-chart
VARIABLES
- Determine the mean, UCL and LCL.
- Estimate an arbitrary range (with same interval).
- Draw the lines.
- Plot the data.
– Number of sample (X-axis)
– Sample Weight (Y-axis) - Interpret the results.
– Out-of-control signals
Construct a Control Chart
– p-chart
– c- chart
ATTRIBUTES
establishes a high degree of certainty
Validation
Validation Parameters
- Accuracy
- Precision
- Selectivity
- Linearity
- Range
- Sensitivity
- LOD
- LOQ
- Ruggedness
- Robustness
– Measure of exactness
Accuracy
Measure of degree of “reproducibility/repeatability” under [normal] condition
Precision
– Ability to measure the analyte of interest
Selectivity/Specificity
Ability to produce results that are directly proportional to concentration of analyte
Linearity
Lowest concentration of analyte that can be detected/measured
Limit of Detection (LOD) / Limit of Quantification (LOQ)
Highest and lowest level of analyte that can be measured
Range
Degree of reproducibility under variety of test conditions
Ruggedness
– Ability to remain unaffected by small variations
Robustness
– Ability to record small variations (changes)
Sensitivity
Meaning of MIL-STD
Military Standard
