|
| |||||||||||||||||||||
|
Om
IVFPERFORMANCE-BASED QUALITY ASSURANCE OF ELECTRONIC
HARDWARE
With a Focus on Shortening the Time-to-Market
Per-Erik Tegehall, Ph D
per-erik.tegehall@ivf.se
IVF Industrial Research and
Development Corporation
Order
#
V040098
Price
SEK 400,00
[+ VAT and
postage)
On behalf of VI (The
Association of Swedish Engineering Industries), IVF has written a report
that describes the traditional practice for reliability assurance, changes
that have occurred that render this approach inadequate, and the
development towards a more adequate methodology for reliability assurance
taking place around the world and then especially in the US. The report
deals with reliability as a part in the context of assuring the quality of
a product.
For Summary,
Contents,
and Introduction
see below. To order the publication fill out the form here
»
Summary
The traditional practise to assure quality of electronic hardware is to design and produce according to standards. These standards regulate design and manufacturing processes but there are also standards that in detail regulate qualification procedures and requirements of the materials and components used for producing electronic products. The requirements in the standards are based on the experience gained from extended use of the materials, components, and processes in question. This experience-based approach has without doubt been tremendously successful enabling also small companies to produce reliable electronics in a cost-efficient way. However, during the 1990s with its fast introduction of new technologies it has become evident that this approach is a dead end. The main objection against this approach is that it cannot assure manufacturability and reliability when immature technology is used. Therefore, any company utilising new technology risk ending up with a costly manufacturing process and delays in product release due to manufacturability and reliability problems.
Another drawback with the standards-based approach for assuring quality is the lack of a clear definition of who has the responsibility for a product’s quality. Logically, it should be the one who formulates the requirements and manufacturing instructions that has this responsibility. In the standards-based approach, this is carried out by the standardisation bodies. However, the standardisation bodies cannot be held responsible for bad quality due to use of their standards.
For these reasons, a new approach needs to be developed in which the responsibilities of all involved parties are made clear. The development of a new approach with these objectives has been initiated by U.S. military authorities and has then been driven by U.S. Army Material Systems Analysis Activity and CALCE Electronic Products and Systems Center at University of Maryland. They have worked closely with the IEEE Reliability Society and their work has resulted in a Reliability Program Standard, IEEE P1332.
In the IEEE standard, the responsibility for reliability is formulated in three statements. It is the responsibility of the supplier to:
-
assure that he has understood the requirements of the customer,
-
develop a process that will satisfy the requirements,
-
adequately verify that the requirements have been met.
It is also stated in the standard that the supplier has the responsibility and the freedom to determine how this shall be achieved. Neither the standard nor the customer shall specify the tasks to be performed.
The statements in the IEEE standard may seem deceptively trivial but they imply a revolution in the way reliability is assured. This new approach has a strong focus on the performance of the product, and can therefore be termed performance-based quality assurance. The aim of the performance-based approach is to proactively incorporate quality into the design process by establishing a scientific basis for evaluating new materials, components, structures, and manufacturing technologies. To achieve that, knowledge of the root causes of failures (physics-of-failure) is essential.
The IEEE standard focuses on reliability issues. However, in addition to reliability, many other issues determine the quality of a product. Furthermore, there is a strong pressure on companies to reduce the time-to-market. Therefore, performance-based quality assurance must cover all important quality issues and facilitate that new products can be introduced to the market in as short a time as possible. To achieve that, the work must be focused on qualifying the design and manufacturing processes used for producing a product rather than the end product itself.
A transition to a performance-based approach is not possible to achieve overnight. It has to be done in steps. It is important that the consequences of performance-based quality management are fully understood before it is adopted in order to assure that adequate resources are allocated. A performance-based approach will mean that the design process will take somewhat longer time and carry a higher cost, but by reducing manufacturability and reliability problems the time-to-market will be shortened and the lifetime costs will be reduced.
The performance-based approach requires extensive knowledge of materials, build-up of components and printed boards, manufacturing processes, field environments, failure mechanisms, reliability tests and screens, simulation tools, and failure analyses. The costs for continuously updating this knowledge and developing new more adequate test methods and simulation tools are very high. Furthermore, performance-based quality management also requires effective systems for communication and handling of information. All data need to be in a form that facilitates its efficient use. Therefore, the most probable progress is that some companies or organisations will take the development of the required knowledge as their core business. This will also mean that the knowledge and tools necessary for an effective adoption of performance-based quality assurance will not be openly available.
This report discusses the consequences of and the changes necessary for a transition to a performance-based approach.
Top
Contents
2.1 Standards-Based Quality
Management.
4
2.2 Performance-Based Quality
Management.
5
2.3 Quality versus
Reliability.
6
3
Current Practice for
Assurance of Quality..
9
3.1 Military Methodology – Basic
Philosophy.
9
3.1.3
Reliability Specification, Allocation, Modelling, and Prediction.
11
3.1.4
Reliability Engineering Design Guidelines.
15
3.1.5
Reliability Data Collection and Analysis, Demonstration, and Growth.
17
3.1.6
Reliability Engineering during Production and Use.
18
3.2 Application in Practice –
Military.
20
3.3 Application in Practice –
Commercial.
21
4
Factors making Current
Practice Inadequate for Assuring Reliability and
Manufacturability..
24
4.1 MTBF-Values are not Adequate for
Predicting Reliability.
25
4.1.1
Prediction of Product Reliability in a Test Chamber or in the Field.
25
4.1.2
Reliability Comparison of Competing Designs.
26
4.1.3
Reliability Design Guidelines.
27
4.2 Failures in Electronic Systems are
not Mainly Caused by Component Defects.
27
4.2.1
Defects at Board Level
28
4.2.2
Wearout Failure Mechanisms.
28
4.2.3
Impact of Handling, Processing and Field Environment
29
4.3 The High-Reliability Market is no
Longer the Driving Force for Development of New Technology.
30
4.5 The Acquisition Process Needs to be
Changed.
33
4.5.1
U.S. Military Acquisition Reform..
33
4.6 The Old Type of Standards will
Vanish.
36
4.7 Changes Needed in a New
Approach.
37
5
Quality Management
Systems.
39
5.4 Conclusions about Quality
Management Systems.
45
6
New Methodology Founded on
Performance-Based Quality Management
47
6.1 Activity 1: Product Requirements
and Constraints.
49
6.2 Activity 2: Product Life Cycle
Conditions.
49
6.4 Activity 4: Qualification of
Packaging Concepts and Manufacturing Processes.
51
6.4.4
Environmental Compatibility.
58
6.5 Activity 5: Risk Management and
Balance of Functionality, Quality, and Cost Requirements.
58
6.5.1
Risk Management of Supplied Materials and Parts.
59
6.5.2
Risk Management of Manufacturing Processes and New Technologies.
59
6.5.3
Failure Modes and Effects Analysis.
59
6.6.2
Defects Caused by Manufacturing Processes.
62
6.7 Activity 7: Failure Analysis and
Feedback of Gained Knowledge.
63
7
Changes Required for
Implementing Performance-Based Quality Management
64
7.1 Concurrent Engineering and the Role
of Management.
64
7.2 Systems for Effective Communication
and Handling of Information.
65
7.3 New Participants in the Supply
Chain Focusing on Quality Assurance.
65
7.4 Research Needs to be Carried out in
Sweden and Europe.
65
APPENDIX A: U.S. MILITARY STANDARDS
Top
Introduction
Production of electronics has gone through dramatic changes on many levels during the last decade. The life cycles of products are becoming shorter and shorter. This development forces companies to revise their product realisation process in order to make it more effective. A delay in the release of a new product is likely to cause decreased market shares or even a completely lost market if the delay is too long [1].
Not long ago, most Original Equipment Manufacturers (OEMs) were vertically integrated doing almost everything from chip to software to final system. The increased global competition and the focus on time-to-market and costs (time-to-profit) have forced the OEMs to spin off or sell off their non-core businesses in order to stay competitive. Extremely successful companies such as Microsoft and Intel have demonstrated the validity of adopting horizontal solutions.
Due to the horizontal integration, supply chains are formed for the manufacture of electronic systems where the companies in the chain are specializing in their respective core competencies. For example, the participants in the supply chain for avionics may include electronic parts manufacturers (components, printed boards, etc.), distributors, contract manufacturers (printed board assembly), system designers, subsystem manufacturers, system manufacturers (e.g. flight control system), customers (aircraft manufacturer), and users (airline) [2].
The focus on the companies’ core competencies has not only made it possible for the companies to shorten the time-to-profit [2]. Spin-off companies have become free to innovate and widen their product offerings. This has lead to an increase in available parts but also to an increased number of sources from which to select parts. Furthermore, it promotes development of parts with improved performance that are smaller, lighter, and less costly at an unprecedented rate. This is one of the main factors behind the fast progress of the electronics industry. However, at the same time as it offers vast opportunities, it is also a big threat to companies. Choosing the right packaging concept may be crucial for the success of a product or even, in some cases, for the survival of the company. With the ever-increasing number of available packaging concepts, this is becoming a true challenge for companies.
There are also other costs to be paid for this development toward horizontal integration. Since each participant in the supply chain will try to optimise his profit, the major players will focus on lucrative high-volume products. Companies outside the main stream will experience increased difficulties in finding suitable suppliers covering their specific needs. Before the trend toward horizontal integration started, it was the high-reliability market that was the driving force for development of new technologies. Since then, a shift has occurred toward the markets for consumer electronics and information technology products. Today, less than 0.25 % of components are used in military products [2]. As a consequence, most of the major component manufacturers have stopped producing components for the military market. It is probably just a matter of time before many of them also exit other high-reliability markets such as automotive, industrial and some telecom applications. The shift toward the consumer market has also had the consequence that many of the newly developed packaging concepts have considerably lower reliability. This development confronts manufacturers of high-reliability products with a new situation that forces them to find new ways to assure the quality of their products.
Even if the component manufacturers continue to produce components qualified for the high-reliability market, the participants in the supply chains for both high-reliability and consumer products would still be challenged to find new ways to assure quality. Quality is today mainly assured by the use of standards developed by U.S. military authorities and other standardisation bodies. These standards systems have still basically the same approach as they had when the first standards were developed about fifty years ago. In short, they assure quality by regulating in detail how to produce electronics and how to qualify materials and components used for producing electronics. The requirements are based on the experience gained from use of the materials, components, and processes that are regulated. This standards-based approach has without doubt been very successful but since the standards are experience-based, i.e. based on “yesterdays” technology, they may not be relevant when using new technology. A company has two alternative ways to handle this. It can either take the risk to get manufacturability and/or reliability problems when using new technology or be conservative and avoid new technologies, which will cause that the use of new technology is discriminated and hampered. In both cases, it may result in products that are not competitive.
Another drawback with this standards-based approach to assure quality is that it becomes very unclear who has the responsibility for a product’s quality. Logically, it should be the one who formulates requirements and manufacturing instructions. In the standards-based approach, this is carried out by the standardisation bodies. Remarkably, the situation that use of standards results in bad quality although they have been followed in every detail is not dealt with in standards, as if it has not even been considered that it could occur. So who is responsible for bad quality in such a case? Probably no one would even think of suing the standardisation body. It is understood that anyone that uses a standard has the responsibility for assuring that it is relevant for that specific application. That would then mean that a buyer that specifies that a product should be manufactured according to specified standards would be responsible for bad quality caused by the use of the standards. A supplier would then only be responsible for assuring that the standards are followed in every detail. That is perhaps convenient for the supplier but not a good way to assure quality. The supplier is the one (or should be the one) that has the best knowledge of the design and/or manufacturing processes he uses and should therefore be the one that has the responsibility for the quality of what he supplies.
During the 1990s, it has become apparent that the present standards-based approach for assuring quality has come to a dead end. The U.S. military authorities, that once lead the development of the standards-based approach, have taken the lead to install a new approach. In 1994, the U.S. Department of Defense (DoD) decided to cancel many of the military standards without replacement and they also declared that acquisitions in the future instead should be based on performance terms. Other standardisation bodies have also declared that they will focus on end-product performance.
Many companies are reluctant to adopt this new performance-based approach (many companies have probably not even heard of it). One reason for this reluctance is probably that they have not fully understood the consequences of the approach and therefore are awaiting a new set of standards that will replace the old standards, standards that will tell them in detail what to do in a similar manner as the old standards did. That will never happen. There will be guidelines but these will only give advice of what needs to be done. Exactly how it should be done is up to the supplier to determine.
The consequences of this new performance-based approach are vast and some consequences are still difficult to foresee. Remember that this transition has to take place at the same time as there is a strong pressure to shorten the time-to-market. Therefore, the transition requires a very well-structured, thought-through strategy where the activities, the roles and responsibilities of everyone involved are clearly specified. A supporting infrastructure must be developed and adequate resources must be allocated. It will require organisational changes and staff with new competencies. Furthermore, all the participants in the supply chain must be involved in this process if it shall be successful. The present trend of outsourcing increases the number of companies that need to be involved, which will make the transition even more difficult.
It is important that the consequences of performance-based quality management are fully understood before it is adopted in order to assure that adequate resources are allocated. If the work and changes required are underestimated, the transition process may easily be terminated. It is hard to find any literature that describes what actually performance-based quality management means and its full consequences. Usually, only some aspects are described and in a way that diminish what is actually needed for the transition. Therefore, in this report I will try to describe in more detail the consequences of performance-based quality management.
In order to discuss changes necessary for the transition to the performance-based approach, it is imperative that the practice used today for assuring quality and the philosophy it is based on is first fully understood. This will first be described followed by a more thorough scrutiny of the changes that has made this practice inadequate. Then, changes necessary to meet this development in order to assure the manufacturing of quality products will be discussed. Since many companies see quality management systems as the primary tool for assuring quality today, they will be the starting point for this discussion.
The report will cover quality aspects of the development of electronic hardware in general but the main focus will be on how to assure manufacturability and reliability on assembly level.
|
|
| Ansvarig: Dag Andersson dag.andersson@ivf.se |
| © IVF Ändrad senast 2002-09-16 |