Sampling by Variables (ANSI/ASQ Z1.9)
OVERVIEW
ANSI/ASQ Z1.9 (formerly MIL STD 414) uses sample statistics to
determine whether a production lot, or incoming lot, should be accepted or
rejected. These statistics also allow estimation of the nonconforming fraction
at each specification limit. Attendees will learn how to define a sample plan
on the basis of (1) the lot size, (2) the inspection level, and (3) the
acceptable quality limit (AQL).
ANSI/ASQ Z1.9 can be used for single specification limits,
two-sided specification limits, and even two-sided limits with different AQLs.
Attendees will learn everything they need to know to address all three of these
applications.
Switching rules meanwhile dictate normal, tightened, or reduced
inspection according to production performance. Poor quality results in
tightened inspection, while a consistent history of good quality and consistent
production conditions allows reduced inspection. In addition, if the process is
under control and its standard deviation is known, a considerably smaller
sample size is required.
 |
| Acceptance Sampling Alternatives |
WHY SHOULD YOU ATTEND
ANSI/ASQ Z1.9 is a widely recognized and generally accepted
standard for acceptance sampling by variables. It uses the average and standard
deviation (or range) of a sample of n parts to determine whether the lot from
which the sample was drawn should be accepted or rejected. The technology is
very straightforward, but users must also be aware of the often-overlooked fact
that the standard relies on the assumption that the part measurements follow
the normal or bell curve distribution. If this is not the case, the standard
will not deliver the intended or expected results.
Attendees will also learn pitfalls that are involved in use of
the sampling tables, and also use of the range method. It is important to
remember that this approach uses the average range of samples of 5, rather than
the range of (for example) all 15 parts in the sample.
Handouts will include a technical appendix that shows the
derivation of some of the material in the standard such as operating
characteristic (OC) curves, and an Excel spreadsheet that duplicates tables B5
and D5, which return estimates of the nonconforming fraction in the lot when
the standard deviation is estimated from the sample statistics (B5) or known
from prior experience (D5).
AREAS COVERED
- Sampling by variables (ANSI/ASQ Z1.9) requires a far
smaller sample than sampling by attributes (ANSI/ASQ Z1.4), but it also
requires real number measurements that may be harder to obtain than pass/fail
data.
- The sample plan, in terms of sample size n and the
acceptance criteria, is a function of (1) the lot size, (2) the inspection
level, and (3) the acceptable quality level (AQL). It is important to pay
attention to the "chutes and ladders" aspect of some of the
tables, in which the intersection of the sample code letter and AQL may
contain an arrow that goes to a different row, and therefore a smaller or
larger sample size.
- The decision as to whether to accept the lot depends on
the sample statistics including its average and also (1) the sample
standard deviation, (2) the average range, or (3) the known process
standard deviation. If the latter is available, a considerably smaller
sample is required. When the range method is used, a sample of n (except
3, 4, or 7) is broken down into groups of 5 whose average range is then
calculated.
- The sample statistics can also be used to estimate the
nonconforming fraction at each specification limit, and this approach must
be used in the more complex applications.
- Switching rules dictate whether inspection must be at
the normal, tightened, or reduced level.
- The standard relies on the assumption that the
measurements follow the normal or bell curve distribution. The normal
probability plot, histogram, and other tests can be used to assess this
assumption. If the process standard deviation is known from a process
capability study, then these tests should have already been performed.
LEARNING OBJECTIVES
This webinar will equip attendees to understand and apply the
ANSI/ASQ Z1.9 (formerly MIL STD 414) standard for acceptance sampling by
variables (real number data), along with vital considerations such as the need
to test the measurements for conformance to the normal (bell curve)
distribution.
WHO WILL BENEFIT
Quality engineers, technicians, and inspectors who use ANSI/ASQ
Z1.9 (formerly MIL-STD 414)
For more detail please click on this below link:
Email: support@trainingdoyens.com
Toll Free: +1-888-300-8494
Tel: +1-720-996-1616
Fax: +1-888-909-1882
Comments
Post a Comment