| Top : Enviromental Testing Information : Environmental Testing Seminner : Environmental testing and non-destructive failure analysis technology |
| |
 |
| Environmental testing and non-destructive failure analysis technology |
|
|
|
| |
 |
 |
|
| |
 |
Environmental testing and non-destructive failure analysis technology |
 |
 |
|
| |
|
Shigeharu Yamamoto*
This report will discuss failure analysis, and the methods of non-destructive failure analysis in particular, as used to improve quality by analyzing the occurrence of failures and providing feedback to the departments concerned. The report will also include some examples of testing and analysis.
|
 |
 |
|
|
| |
|
|
|
|
| 2.Approaches of failure
analysis |
|
|
|
|
|
|
| (1) |
Even large systems are composed
of units and individual subsystems, and those
in turn are composed of individual parts,
materials, and components that can be analyzed.
Failures of these individual parts and materials
cause an expanding ripple effect leading to
failure of the entire system and causing major
accidents. Because of this, reliability is
maintained and safety is assured by specifying
the causes of these failures and taking countermeasures. |
| (2) |
The bigger the system, the
more crucial the problems posed by the interfaces
between complex parts and units in which hard
sections of individual parts and materials
are combined with parts and materials of differing
qualities. Accidents due to human contributions
such as misuse and incorrect operation also
form part of this equation. This can be well
understood by looking at the recent Donen
accident and airplane accidents. |
| (3) |
In general the cause of
failure is said from experience to be due
to:
1. Something that can be handled in design
or at the design stage, 80%
2. Something caused in a production process,
15%
3. Something due to usage conditions, 5% Examples
of matters that should be handled at the design
stage include (a) constructions that lead
to mistakes because the items is difficult
to use, cannot be operated easily, or has
complex operation, (b) constructions with
difficult maintenance such as operations or
processes that lead to incorrect operation.
Problems at the design stage are more often
due to insufficient design investigation (time/frequency,
technology, properly qualified personnel)
rather than the design itself. |
| (4) |
When failure is very slight,
such as only one sample item or one set, product
failure analysis looks at whether that failure
is likely to have a major impact in the field.
Waiting to see whether such failures occur
in the field is thought to be too late, and
so nowadays quality such as parts, materials,
processes, and ease of operation are checked
before shipping products to market. A number
of examples are used to illustrate methods
for looking in advance at the possibility
of incorrect operation or failure, such as
advance analysis (good product analysis) in
addition to design investigation. One example
is the pre-analysis of the inner conductivity
conditions of multi-layer PCBs. (Refer to
example 6-1.) Failure analysis considers potential
and actual failure mechanisms and performs
investigative research to provide corrective
measures. These corrective measures can also
be applied to other situations. When a failure
occurs due to some cause, that cause can be
identified as physical, chemical, mechanical,
or electrical, or as due to human error. The
results can always be explained in terms of
the failure being caused according to a particular
development. |
|
|
|
|
|
| 3.Methods of failure analysis |
|
|
|
|
|
|
Failure analysis can
be broadly divided into de-structive and non-destructive
testing. Potential defects and actual failures are
identified and confirmed, and electrical, physical,
chemical, and human engineering investigations are
undertaken.
These activities are crucial to creating
countermeasures for the causes of failure. Fig.
2 shows the general procedure for failure analysis.
|
Fig. 2
Example of failure analysis
|
|
|
|
Two methods of locating failure
are:
(1) The overall method {epidemiological method:
non-destructive testing}, and
(2) The individual method {epidemiological method
(physical failure method): destructive testing}.
For procedure, less time
is lost by taking up the epidemiological method
first and the physical failure method later. An
unexpectedly large number of failures can be found
with the epidemiological method, but when the epidemiological
and epidemiological methods are skillfully combined,
reliability can be greatly improved.
(1) The epidemiological method
This approach uses statistics to explain
the degree of failure (e.g., according to time,
place of occurrence, market distribution, market
location, distribution of usage conditions, distribution
of users, by lot, and time differences in manufacturing
methods). Even if the reason for this type of failure
is not understood, the range encompassing the failures
can be narrowed down, and has the know-how to come
up with effective countermeasures.
This method can be used by anyone with
no need for expensive equipment.
(2) The epidemiological method (physical failure
method)
This approach
relies on completely grasping the mechanisms and
principles of the failure and narrowing the actual
cause to one point. In other words, this method
has the know-how to theoretically understand why
the failure occurred.
Failure analysis clarifies the cause
of failure, understands the failure mechanism, and
links these to preventing recurrence of failure.
In failure analysis, it is vital to
obtain as much information as possible externally
before making a destructive investigation, and normally
non-destructive testing is done first. The extent
of information obtained from this non-destructive
testing will greatly influence later analysis and
countermeasures.
When performing failure analysis, the
following specific items must be considered.
| 1) |
Don't analyze with insufficient
investigation into factors such as shipping
conditions, usage con-ditions, and usage environmental
conditions. |
| 2) |
Don't handle failure samples
carelessly. |
| 3) |
When investigating appearance,
look at such conditions as dimensions, thickness,
weight, shape, bending, color changes, rust,
foreign matter adherence, mold, cracking,
scratching, cloudiness, scorching, and nicking.
|
| 4) |
Since destructive analysis
will cause the original shape and conditions
to be lost, be sure to take plenty of pictures
at the initial stage from every angle, of
the piece as a whole, and with enlargements. |
| 5) |
Carefully observe and record
each time. |
| 6) |
Since failure analysis uses
tools and chemicals, consider worker safety. |
| 7) |
Analyze failure objectively.
Don't become convinced that a particular idea
must be true. |
| 8) |
Since the human factor is
an important cause, consider the background
carefully. |
| 9) |
Analysis relies greatly on
experience, skill, and power of observation. |
| 10) |
It is dangerous to ignore
multiple causes and become convinced of a
single cause. |
| 11) |
Don't destroy the object
by overloading. |
| 12) |
Compare failed items with
good items, and analyze the failures. |
| 13) |
Consider effects from the
workers themselves and from jigs, tools, and
equipment used. |
| 14) |
Don't throw away the specimens
before final results have been attained. Keep
them. |
| 15) |
When making judgements, don't
simply use the data alone as if it were analysis
results. |
| 16) |
Prepare good items from the
same lot as the failed items. When the cause
of failure is suspected to run across lots,
good items must be investigated together across
lots. |
| 17) |
As a rule, proceed from the
general situation to details to the overall
picture. Also, go from external to internal
analysis, from peripheral to the object, and
from non-destructive to destructive. |
| 18) |
Make full use of such elements
as knowledge, ex-perience, and data. Ask for
assistance from another specialist. |
| 19) |
In failure analysis, when
you have gone on to the next process, you
can't go back, so you must observe very carefully
at each stage of analysis. |
|
|
| |
|
|
