 |
 |
|
| |
 |
Guidance for accelerated testing and reliability
- For electrical and electronic parts and equipment - |
 |
 |
|
| |
|
Toshio Yamamoto*
Resources are continually being poured into efforts to improve the reliability of electrical and electronic parts and equipment. This has lead to the production of a large number of highly reliable products, and brought about a sharp decline in failure rates. Concretely evaluating product reliability requires extremely long periods of time of environmental testing, specifically in the area of reliability testing. These time constraints have created a need for accelerated testing that can produce effective results in shorter periods of time. However, accelerated testing cannot be achieved by merely ratcheting up the stress to shorten testing times, and this type of testing is not applicable to every situation. Accelerated testing, just like other forms of reliability testing, is a reliable test method supported by scientific logic. In this report, we would like to present the logic behind accelerated testing in an easily understandable way, from the standpoint of environmental testing used in the fields of electricity and electronics.
|
 |
 |
|
|
| |
|
|
| |
| The details of environmental testing differ according to the type of product, and no one test method can serve for all products. We could go so far as to say that each individual product requires its own test procedure. Within the category of reliability testing, environmental testing normally focuses on determining whether a product can fully display and maintain its established functions within the marketplace environment of actual use after the product has been shipped from the factory. This testing is done from the standpoint of technology directly related to the product. However, even the same types of products face a variety of environments in the marketplace, including how the individual product may be used. Because products meet such conditions, each type of environmental test has its own significance. However, accelerated testing is not created by merely increasing the stress in a rush for results. Such methods will simply create destruction testing. Because of this, we must thoroughly understand the principles of accelerated testing, carefully construct the logic involved, and create a method in which no contradictions arise in the testing procedure.
|
|
| |
|
 |
| 2.Definitions, prerequisites, and fundamentals |
|
|
|
|
|
|
| |
According to " JIS (Japanese Industrial Standard) Z 8115 Glossary of terms used in reliability", accelerated testing refers to "a test in which the applied stress level is chosen to exceed that stated in the reference conditions in order to shorten the time required to observe the stress response of the item".In general, the more severe the conditions in relation to standard conditions, the more quickly it is possible to complete the test. However, it would be premature to conclude that we have proven the excellence of a product simply because it has endured an accelerated test. A product that has successfully endured extremely severe conditions may fail in a relatively short time under the same type of stress conditions that are considered extremely mild. We must create a test plan for all methods of use and surrounding environments that can be imagined, and we must carry out complete pre-test studies and check for misconceptions in logic, and make sure we haven't missed anything.
To start with, there is always a cause for each type of product failure. Unless there has been severe misuse, such failures do not merely appear out of the blue. Seen from a qualitative standpoint, the failure of electrical and electronic products, and electronic equipment in particular, begins when changes are experienced in the physical and chemical characteristics of the materials used in their production. Such changes can be caused by an intrusion from the external environment, or from internal factors, or from a combination of both internal and external factors. These physical and chemical reactions are affected by the surrounding environment (heat, in particular), speeding the reaction, and the product functions are weakened are weakened or completely stopped in the course of achieving equilibrium in that reaction. Therefore, accelerated testing consists of speeding up this physical and chemical process with suitable ambient environments, causing malfunctions to appear.
Accelerated testing characteristically entails knowing in advance the results of testing when deciding to run the tests, because one must establish what is being accelerated in relation to what, and clearly establish the items that serve as the basis for acceleration. Essentially, accelerated testing is not a means for ferreting out new modes of failure, but rather serves as a time-saving means of confirming phenomena that have already been identified. Therefore, major premises have already been proposed concerning the factors involved in failure and the mechanism leading to failure. Only when expected results are obtained do the conclusions form accelerated testing. Then, by recording the test preparations and accumulating the data, it finally becomes possible to apply the procedures to similar products. However, there are no accelerated tests that can be applied indis-criminately to all products and failure modes. For example, if the relation between the failure mode and the acceleration mechanism is not known and we just expose a new product to severe conditions and cause the product to fail, the only result will be that we have destroyed the product. We will not be able to obtain any scientifically valid data at all from such testing. We must remember that accelerated testing, as one form of reliability testing, is a very specific technical method.
To sum up, accelerated testing consists of, "acceler-ating the mechanisms causing product failure by using conditions that are more severe than standard conditions, and saving testing time by utilizing the systematic nature existing between the standard conditions and the accelerated conditions in the identical mode of failure". In other words, the absolute prerequisites to accelerated testing are:
(1) Having the same product failure mode, and
(2) Having a systematic relationship between standard conditions and accelerated conditions.
Care must also be taken in the following matters when performing accelerated testing.
(1) The test must be based on the main points whose results have been predicted by thoroughly examining the problems.
(2) )The test conditions must be constructed on the basis of thoroughly grasping the usage conditions and the ambient environment of the product.
(3) ) When special prerequisites exist, they must be dealt with, but one should always try to gather universal data that can be applied in the future.
The most important condition of all is to clarify exactly the purpose for running the test. In other words:
(1) The test presumes a correlation with conditions in the field. (Conditions are within the product ratings, and the test aims at forecasting or reproducing failures that occur in the field.)
(2) )The test attempts to detect the characteristic weaknesses of the product. (With the aim of collecting data to discover and improve product weaknesses)
(3) ) The test simply attempts to confirm product durability. (Durability is confirmed outside the product rating.)
The most important condition of all is to clarify exactly the purpose for running the test. In other words:
|
|
| |
|
|
|
| 3.Accelerated testing and the physical chemistry of product failure |
|
|
|
|
|
|
| |
The physical chemistry of product failure in electrical and electronic equipment can be broadly grouped into approximately 30 categories, or more specifically classified into about 70 categories. The mechanisms leading to failure have mostly been elucidated. Therefore, the method of planning accelerated testing or analyzing product failure quite often consists of selecting the appropriate mechanism leading to predicted failure from among these categories, and then verifying the pre-selected or pre-constructed failure model. First of all, to capture the general concept of failure occurrence, the stress-strength model is briefly checked, and then great emphasis is placed on the chemical kinetics model, which is extremely easy to understand for accelerated testing.
| 3-1 |
Stress-strength model
This model explains the process leading to failure
through the relationship between the characteristic
durability of the product and the stress from the ambient
environment. (Failure occurs in the areas in which
deterioration of durability and distribution of stress
overlap.) |
|
|
|
|
| |
|
|
|
|
