Testing is an important issue in any development programme, because it becomes more important and a significant component of the overall development effort when the development risks are high. Designs are not always perfect and designers usually are not able to analyse all the likely causes of failure of their designs in service, accordingly testing is necessary to determine the reliability of the design, but what kind of tests should be applied?
In fact any kind of test can be considered as a part of an integrated test programme, that should include: Functional performance testing, Environmental testing, Reliability and durability testing and Safety testing. The test programme should be managed and planned correctly into the phases of concept creation, design, development, manufacturing and service. where, the implemented tests during each phase, are different from another phases. Testing especially during the development phase is nearly expensive and uncertain in relation to time and other resources required. The effective selection of testing programs (and the incurred costs due to ineffective selection of testing programs) will be noticeable in the future, so a long- term view is significant.
In fact all the test aspects: what to test, how to test, test plans, test implementation, link design analysis tools to test program. requires sufficient expertise in this field to achieve the goals of the test. Testing regimes at some companies rely on expert testing in specific applications to establish an appropriate testing program and then they carry it out.
The main objectives of this project; is to develop an expert system that is able to provide:
1. A list of most appropriate tests for a system /equipment under consideration.
2. Guidelines for an optimal test plan and explaining the way to implement the test.
Application of expert systems in the field of testing programs will help to design a tool that allows the automatic selection of the most appropriate tests for a system /equipment under consideration. Where the required knowledge can be accessed quickly and easily by managers that will aid the quality of decision-making.
Usage environment, test goals, design analysis tools (FMEA, FTA and ET) and some information from literature will be combined to build a decision tool for designing a robust test process. It is also intended that physics of failure principles will be used to aid the identification of failure modes, failure mechanisms and stressors.
Table of Contents
Section (1) Introduction
1.1 Expert systems
1.2 Test engineering
1.3 Design analysis method for reliability
1.4 The proposed expert system for effective selection of testing regimes for engineers (ESFEST) considerations
1.5 Section (1) Conclusion
Section (2) The methodology of building a knowledge base for (ESFEST)
2.1 Introduction to The Methodology
2.2 Physics-of-failure approach
2.3 Failure mechanisms/ Failure modes of Materials /components
2.4 Stressors (accelerated tests)
2.5 Building the Knowledge Base
2.6 Section (2) Conclusion
Section (3) ESFEST, design the Process and building the system
3.1 Introductions to the process
3.2 What ESFEST is?
3.3 The process detail and building the expert system (main diagram)
Section (4) Case study - Electrical motor
4.1 Introduction to case study
4.2 tests selection
Section (5) Conclusion Link ESFEST to CAD system
Research Goals and Themes
This work aims to develop an "Expert System for Effective Selection of Testing Regimes" (ESFEST) to assist engineers in designing robust reliability testing programs by automating the selection of appropriate tests based on product materials, failure modes, and environmental stressors.
- Development of a knowledge base utilizing physics-of-failure principles.
- Integration of design analysis tools like FMEA, FTA, and Event Trees.
- Methodology for mapping failure mechanisms to specific accelerated test stressors.
- System architecture for decision-making in testing regimes for electrical/mechanical systems.
Excerpt from the Book
1.1.2 The structure of expert systems:
An expert system can be considered as four main parts: Induction of knowledge: The first step the user must do is entering the knowledge into the system; the way to do that is related to the level of complication of the expert system. If the expert system is ‘ruled-based system’ the knowledge can be entered into the system as a set of rules, while if the expert system is ‘rule-inducing’ the rules can be extrapolated from existing data or past performance. A rule-inducing package can be stored for future use and interrogated whenever it is necessary. [5]
Where ‘ruled-based system’ is an axiom of artificial intelligence in which intelligent behaviour is ruled –governed. [6]
And ‘Rule-inducing’ in which a system can be used as an existing data or past performance such as data that is held in a database and try to work out if there are any inherent rules in the data, which are valid and consistent. [5]
Chapter Summary
Section (1) Introduction: This chapter provides fundamental definitions of expert systems and test engineering, establishing the context for why an automated selection system is necessary for reliability testing.
Section (2) The methodology of building a knowledge base for (ESFEST): Discusses the logic behind building the knowledge base using physics-of-failure concepts to link failure modes with appropriate stressors and test methods.
Section (3) ESFEST, design the Process and building the system: Details the architecture and operational flow of the expert system, including input identification, process programs, and the interaction between the system and its memory.
Section (4) Case study - Electrical motor: Applies the developed ESFEST methodology to an electric induction motor to verify the practical selection of effective accelerated tests.
Section (5) Conclusion Link ESFEST to CAD system: Explores the integration of the expert system with CAD/CAE software to enable two-way communication between design tools and reliability analysis.
Keywords
Test engineering, Reliability Testing, AI, Expert systems, FMEA, FTA, ET, Accelerated testing, Physics-of-failure, ESFEST, Knowledge base, Failure modes, Stressors, Design analysis, CAD integration
Frequently Asked Questions
What is the core purpose of this research?
The research proposes an expert system (ESFEST) designed to automate and optimize the selection of reliability testing regimes, helping engineers identify the most appropriate tests for their specific product designs.
What are the central thematic areas?
The work focuses on the intersection of artificial intelligence (expert systems), test engineering, reliability modeling (FMEA, FTA, Event Trees), and the physics-of-failure approach to material degradation.
What is the primary objective of the ESFEST system?
The system is meant to provide a list of appropriate testing methods and guidelines for implementing an optimal test plan for a given set of equipment or materials.
Which scientific methodology is primarily employed?
The development utilizes a "physics-of-failure" methodology, which identifies root-cause failure mechanisms based on materials, then maps them to environmental stressors, and finally to valid test treatments.
What topics are covered in the main body (development stages)?
The main body covers knowledge base construction, logical rule-based process programs, integration with design analysis documents, and validation through a case study on an electrical induction motor.
Which keywords best characterize the work?
The core keywords include Test engineering, Reliability Testing, Physics-of-failure, Expert systems, FMEA, FTA, Accelerated testing, and ESFEST.
How does the proposed ESFEST handle failure data?
It uses a Blackboard data structure where independent knowledge bases (KBs) specialized in different domains (materials, failure modes, mechanisms, stressors, tests) interact indirectly to produce recommendations.
What is the significance of the Electrical Motor case study?
The case study serves as a practical validation of the theoretical framework by demonstrating how the ESFEST would navigate the testing requirements for a specific mechanical device exposed to environmental stressors like dust, heat, and vibration.
- Quote paper
- Hosam Aljaz (Author), 2003, Development An Expert System For Designing Effective Reliability Testing Programs, Munich, GRIN Verlag, https://www.grin.com/document/1557467
