Forensic engineering is the investigation of materials, products, structures or components that fail or do not operate/function as intended, causing personal injury or damage to property. The consequences of failure are dealt by the law of product liability. The subject is applied most commonly in civil law cases, although may be of use in criminal law cases. Generally the purpose of a forensic engineering investigation is to locate cause or causes of failure with a view to improve performance or life of a component, or to assist a court in determining the facts of an accident. It can also involve investigation of intellectual property claims, especially patents.
Methods used in forensic investigations include reverse engineering, inspection of witness statements, a working knowledge of current standards, as well as examination of the failed component itself. The fracture surface of a failed product can reveal much information on how the item failed and the loading pattern prior to failure. The study of fracture surfaces is known as fractography. Fatigue often produces a characteristic fracture surface for example, enabling diagnosis to be made of the cause of the failure. The key task in many such investigations is to identify the failure mechanism by examining the failed part using physical and chemical techniques. This activity is sometimes called root cause analysis. Corrosion is another common failure mode needing careful analysis to determine the active agents in the environment which initiated the corrosive attack. Accidents caused by fire are especially challenging owing to the frequent loss of critical evidence, although when halted early enough can usually lead to the cause. Fire investigation is a specialist skill where arson is suspected, but is also important in vehicular accident reconstruction where faulty fuel lines, for example, may be the cause of an accident.
The broken fuel pipe shown at right caused a serious accident when diesel fuel poured out from a van onto the road. A following car skidded and the driver was seriously injured when she collided with an oncoming lorry. Scanning electron microscopy or SEM showed that the nylon connector had fractured by stress corrosion cracking due to a small leak of battery acid. Nylon is susceptible to hydrolysis in contact with sulfuric acid, and only a small leak of acid would have sufficed to start a brittle crack in the injection moulded nylon 6,6 connector by a mechanism known as stress corrosion cracking, or SCC. The crack took about 7 days to grow across the diameter of the tube, hence the van driver should have seen the leak well before the crack grew to a critical size. He did not, therefore resulting in the accident. The fracture surface showed a mainly brittle surface with striations indicating progressive growth of the crack across the diameter of the pipe. Once the crack had penetrated the inner bore, fuel started leaking onto the road. Diesel is especially hazardous on road surfaces because it forms a thin oily film which cannot be seen easily by drivers. It is akin to black ice in lubricity, so skids are common when diesel leaks occur. The insurers of the van driver admitted liability and the injured driver was compensated.
FMEA and fault tree analysis methods also examine product or process failure in a structured and systematic way, in the general context of safety engineering. However, all such techniques rely on accurate reporting of failure rates, and precise identification, of the failure modes involved.
There is some common ground between forensic science and forensic engineering, such as scene of crime and scene of accident analysis, integrity of the evidence and court appearances. Both disciplines make extensive use of optical and scanning electron microscopes, for example. They also share common use of spectroscopy (infra-red, ultra-violet and nuclear magnetic resonance) to examine critical evidence. Radiography using X-rays or neutrons is also very useful in examining thick products for their internal defects before destructive examination is attempted. Often, however, a simple hand lens suffices to reveal the cause of a particular problem.
Trace evidence is sometimes an important factor in reconstructing the sequence of events in an accident. For example, tire burn marks on a road surface can enable vehicle speeds to be estimated, when the brakes were applied and so on. Ladder feet often leave a trace of movement of the ladder during a slipaway, and may show how the accident occurred. When a product fails for no obvious reason, SEM and Energy dispersive X-ray spectroscopy or EDX performed in the microscope can reveal the presence of aggressive chemicals which have left traces on the fracture or adjacent surfaces. Thus an acetal resin water pipe joint suddenly failed and caused substantial damages to a building in which it was situated. Analysis of the joint showed traces of chlorine, indicating a stress corrosion cracking failure mode. The failed fuel pipe junction mentioned above showed traces of sulfur on the fracture surface from the sulfuric acid which had initiated the crack.
Most manufacturing models will have a forensic component that monitors early failures to improve quality or efficiencies. Insurance companies use forensic engineers to prove liability or alternatively non liability. Most engineering disasters (structural failures such as bridge and building collapses) are subject to forensic investigation by engineers experienced in forensic methods of investigation. Rail crashes, aviation accidents and some automobile accidents are investigated by forensic engineers particularly where component failure is suspected. Furthermore, appliances, consumer products, medical devices, structures, industrial machinery, and even simple hand tools such as hammers or chisels can warrant investigations upon incidents causing injury or property damages. The failure of medical devices is often safety-critical to the user, so reporting failures and analysing them is particularly important. The environment of the body is complex, and implants must both survive this environment, and not leach potentially toxic impurities. Problems have been reported with breast implants , heart valves, and catheters, for example.
Failures which occur early in the life of a new product are vital information for the manufacturer to improve the product. New product development aims to eliminate defects by testing in the factory before launch, but some may occur during its early life. Testing products to simulate their behaviour in the external environment is a difficult skill, and may involve accelerated life testing for example. The worst kind of defect to occur after launch is a safety-critical defect, a defect which can endanger life or limb. Their discovery usually leads to a product recall or even complete withdrawal of the product from the market. Product defects often follow the bath-tub curve, with high initial failures, a lower rate during regular life, followed by another rise due to wear-out. National standards, such as those of ASTM and the British Standards Institute, and International Standards can help the designer in increasing product integrity.
It is unfortunate that product failures are not more widely published in the academic literature or trade literature, partly because companies do not want to advertise their problems. However, it then denies others the opportunity to improve product design so as to prevent further accidents. However, a notable exception to the reluctance to publish is the journal Engineering Failure Analysis, which publishes case studies of a wide range of different products, failing under different circumstances. There is also an increasing number of textbooks becoming available.
- Introduction to Forensic Engineering (The Forensic Library) by Randall K. Noon, CRC Press (1992).
- Forensic Engineering Investigation by Randall K. Noon, CRC Press (2000).
- Forensic Materials Engineering: Case Studies by Peter Rhys Lewis, Colin Gagg, Ken Reynolds, CRC Press (2004).
- Peter R Lewis and Sarah Hainsworth, Fuel Line Failure from stress corrosion cracking, Engineering Failure Analysis,13 (2006) 946-962.
There is no pharmaceutical or device industry support for this site and we need your viewer supported Donations | Editorial Board | Governance | Licensing | Disclaimers | Avoid Plagiarism | Policies