The Acceptance Quality Limit is a proven statistical instrument for quality control of products.
The Need for Quality Control
Virtually any production batch will contain faulty products, although defects can be very rare. There will always be faulty products, even if a quality check has been carried out and any defects found have been corrected. Perfect products only exist in a perfect world, which is why we strive to keep the deviations of our products within certain limits.
While primary packaging manufacturers’ quality assurance systems are becoming increasingly efficient through the use of innovative technologies such as electronics and laser technology, there will always be room for improvement; every production process offers the opportunity to develop new quality standards.
A buyer always wants perfect products, but the supplier cannot be expected to deliver only perfect goods. However, the buyer wants to check whether the goods purchased are of consistently high quality because he does not want products with (too many) defects.
Background of the AQL
It is impossible to fully check the functionality of each individual product. Products would then become far too expensive or immediately unusable because many test methods destroy the product. That is why the method of “acceptance sampling” has been developed (samples used to check whether the quality of a product is acceptable). Acceptance sampling is a statistical quality check, a middle way between a full AQL inspection.
Acceptance sampling was invented before World War II when the US military had to determine which batches of ammunition were acceptable and which bullets were to be rejected. The tested ammunition was destroyed. The United States military developed the military standard 105-D, testing a pair of representative bullets from a batch of ammunition so that one could be confident that the other bullets would work properly on the battlefield. Ultimately, this could mean the difference between life and death.
The test method in accordance with military standard 105-D has resulted in a statistical sampling method that is used worldwide, the so-called AQL. The AQL tables are static tools that product and quality inspectors can use. Quality assurance inspectors can use the AQL tables to determine which sample size leads to reliable results and which standard to use for acceptance and rejection.
The random sample selected in this way, in which representative samples (with a certain sample frequency) are chosen from a number of products (a lot) and are tested to determine whether a lot of products is acceptable or should be rejected, is called an acceptance plan or sampling plan. This plan must comply with DIN ISO 2859 part 1. The AQL, defined as the lowest quality level that is still accepted, is therefore a statistical procedure that determines the quality of the batch of products.
AQL Standard for Blown Glass
Primary packaging materials made of blown glass for pharmaceutical and cosmetic purposes are often produced in accordance with quality agreements based on the Defect Evaluation List for Containers Made of Molded Glass and the Defect Evaluation List for Containers Made of Molded Glass. relevant WHO GMP rules.
The acceptable number of defects:
To determine the maximum number of defective units (the acceptable number of defective products), a quality inspector must know:
- The size of the batch of products
- What the inspection level is
- The size of the random sample
- What is the so-called AQL share?
To determine the correct sampling plan and level of acceptance/rejection, quality inspectors mainly use two AQL tables:
- The first table determines which ‘code letter’ should be used given the inspection level and the size of the batch of products.
- The second table defines the sample size and the maximum number of defects that can be accepted based on the code letter and AQL.
Classification
The most commonly used classification of blown glass defects and the corresponding AQL shares are as follows:
- Class 1 defect: critical defect; therefore, the AQL share is 0
Result: packaging material cannot be used - Class 2A defect: serious defect; consequently, the AQL share is 0.65
Result: usability of packaging material significantly reduced - Class 2B defect: serious defect; consequently, the AQL share is 1.5
Result: usability of packaging material partially reduced - Class 3 defect: minor defect; consequently, the AQL share of defects is 6.5
Result: usability of packaging material somewhat reduced
Rules for Counting Defects:
For each class of defects, the number of defects is usually determined according to the following method:
- Each individual defect is counted.
- If a bottle has more than one flaw, it will only be counted once.
- If a bottle has more than one flaw from more than one class, the lowest AQL will be applied.
Example of the Assessment of the Number of Defects
During a quality assurance test, an internal open glass bubble (deviation reference number 14.05.11) was found on two batches of 550,000 blown glass bottles. According to the current Defect Evaluation List, this is a class 2A defect, for which an AQL share of 0.65 applies. What are the correct sample size and an acceptable number of defects?
A Batch of Products from 550,000 Blown Glass Bottles.
- The inspection level is determined to be General inspection level ‘II’. The code letter is, therefore, “Q”.
- Consequently, the size of the random sample from the total batch of products is 1 250 units.
- Class 2A defect: serious defect; consequently, the AQL share of defects is 0.65
- The acceptable number of bottles with defects is a maximum of 14 with an AQL share of 0.65
- The unacceptable number of defective bottles is 15 or more with an AQL share of 0.65
Results of Counting at Lot 1
The quality inspectors found the following numbers of bottles with defects:
- 12 bottles with various defects of class 2A
- 10 bottles with various defects of class 2B
- 12 defects of class 2A are less than the highest acceptable number of defects (14).
- 10 defects of class 2B is less than the highest acceptable number of defects.
Result: A batch of products 1 is released/approved, as none of the limits for the two classes of defects (2A and 2B) are exceeded.
Results of Counting in Game 2
The quality inspectors found the following numbers of bottles with defects:
- 14 bottles with various defects of class 2A
- 22 bottles with various defects of class 2B
- 14 defects of class 2A equal the highest acceptable number of defects (14).
- 22 defects of class 2B is more than the highest acceptable number of defects (21, see table 2).
Result: A batch of product 2 is rejected/rejected because one of the limits for the two classes of defects (2A and 2B) is exceeded.
The examples are given and the applied counting rules and AQL shares are for illustrative purposes only.
Interesting Related Article: “5 Ways to Reduce Software Development Costs without Losing Quality“
from Business – Market Business News https://ift.tt/34I2uek
via IFTTT
0 Comments