ABSTRACT
The techniques used in the investigation of engineering failures are wide ranging and are continuously expanding as more and more sophisticated instruments, equipment, and computing software become invented and made affordable. These are used, for example, for the measurement of miniscule amounts of elements, or for the examination of sub-microscopic phases or surfaces of materials, or for the 3-dimensional modeling of stresses and deformations of complex objects. All these have allowed the Failure Investigator (FI) to identify the proximate and root causes of a failure more accurately than ever before. At the same time, parallel developments in other elds have enabled the FI to produce reports faster. For example, with the present developments in digital photography and printing, computer hardware and software, and Internet technology, the FI is able conduct site examination, process the relevant photographs, look up some online literature, write a short preliminary report, and have a hard copy ready by the next day, or have an electronic copy sent by Internet halfway round the world in an even shorter space of time. Two decades ago, such speed and efciency was quite inconceivable. But, despite all of the above advancements, a carefully planned procedure of investigation, directed by human skills and performed by human investigators, must rst be conducted before any ndings can be made or any report can be produced. Such a procedure should include some or all of the following steps:
1. Obtaining background information 2. Physical examination/testing and chemical analysis 3. Stress analysis and computer modeling 4. Report writing
In the process of conducting failure analysis (FA), the FI has to look beyond the normal limits of design methodology, and identify the actual conditions causing the failure, and sometimes check the design. This requires
familiarity with the major types of stresses and strains and failure modes, disciplines that can normally only be obtained through a rigorous academic program, experience on the eld, and self-study. The FI also needs to be recognized as being professionally competent by the regulations in force, and to be deemed an expert witness, whose ndings can be accepted in court. The FI is required to be a technical detective as well as a student of human psychology, for the simple reason that persons who have been negligent and have caused the failure do not normally incriminate themselves by telling the truth. The FI should be wise enough to detect an untruthful witness.
