An investigation grounded in sound scientific data can ascertain the mode and root cause of component failure with minimal time and money.
|
Figure 1: Severe cracking was apparent through the chamber outlet nipple. Photo: Stork Technimet Technicians.
|
Premature failure of manufactured components is a continual problem in today’s high-paced manufacturing environment, in which emphasis is placed on taking a product rapidly to market and manufacturing that product quickly and cost effectively. Product failure can result in a wide range of consequences with far reaching financial impact. Negative outcomes include the expense of replacement parts and the often greater cost of installation; legal fees resulting from lawsuits arising out of property damage, physical injury, and death; and a potential decay in product reputation and the subsequent loss of revenue. Often the response to a failure is a rush to correct the problem without an eye toward addressing the root cause. By conducting a thorough investigation and failure analysis, science can be applied to a negative situation and benefits can actually be derived.
Conducting a failure analysis is analogous to piecing together a jigsaw puzzle. In both cases, the end result is the gathering of many individual bits, which by themselves are often meaningless, but together create a complete picture or tell a full story. Failure analysis is the application of analytical observation and testing—guided by engineering practices—to the investigation of a failure within a part or assembly.
The obvious purpose of such an investigation is to gain a comprehensive understanding of the failure in order to make immediate corrections in existing components or manufacturing processes. However, the results of a complete failure analysis can have more far-reaching benefits.
Such an investigation can generate data that will allow improvements or modifications in future parts. The information provided can assist with changes in product design, material selection, production and assembly methods, and inspection and evaluation procedures, through the identification of systemic defects or flaws in a product.
A failure analysis can be performed at any point in the product life cycle. Products that fail during engineering evaluations are routinely analyzed as a means to improve the component. Failures occurring while in service are commonly investigated to assess whether aspects of the product or the service environment were responsible. The saying, “If you don’t know how it broke, you don’t know how to fix it”, expresses the significance of carrying out a formal failure analysis.
In spite of the obvious benefits of performing a formal failure analysis, often this process is overlooked in favor of pursuing pet theories. Sometimes this is justified based on the both the time required to conduct the evaluation, and the cost of such an investigation. A study reported by Michael Sepe illustrates that problem solving through analytical testing guided by sound engineering practice can save many times the actual cost of the investigation, and reduce the overall time required to solve the problem by weeks (Injection Molding, September, 2001).
The objectives of a failure analysis are to ascertain the mode and the cause of the component failure; in other words to determine how the part failed and why it failed. In cases of cracking with plastic components, failures can occur through many different mechanisms, including catastrophic ductile overload, creep, environmental stress cracking, molecular degradation, and fatigue. Determining the failure mode involves identifying how the fracture initiated, and how the resulting crack extended.
How and why
Assessing the mode of failure is frequently easier than determining why the part failed. Often a single cause cannot be identified, with multiple factors contributing to the failure. All of the factors affecting the performance of a component can be placed into one of four categories, those being: material, design, processing, and service conditions. It would be incorrect to think that these factors act independently on the component; it is the combination of these factors that determines the performance properties of a component.
Conducting a proper failure investigation requires an unbiased approach, grounded in sound scientific data collection, without hidden agendas or ulterior motives. This is particularly evident when outside suppliers are involved, and when accusations of fault and liability are made. Preconceptions regarding the failure need to be set aside, and the evaluation needs to be carried out with an open mind. This is not to say that common sense and historical perspective should be ignored. On the contrary, these are valuable tools in a failure investigation. However, the facts should guide the investigation and the interpretation of the obtained data. In many cases it may be advantageous for an independent test laboratory to perform the analysis, as such an organization may be in the best position to be impartial.
A proper failure analysis, like assembling a jigsaw puzzle, should be performed in a systematic manner. This process includes the collection of pertinent information surrounding the component and the failure; a thorough examination of the failed component; analytical testing; and finally a comprehensive review of the obtained data. The process of conducting a failure analysis follows the same basic steps and principles for both plastic and metal components, with the chief differences rooted in the analytical techniques used to evaluate the composition and structure of the respective materials.
|
Figure 2: SEM image showing crack initiation along the inner wall. (Original magnification 40x). Photo: Stork Technimet Technicians.
|
This case study illustrates the process and importance of conducting a failure analysis. Several flow regulator chambers had failed after limited service in a distilled water application. The chambers had been used in conjunction with municipal water service and had been exposed to temperatures approaching 100°C for limited periods. The chambers were molded from a non-reinforced grade of polycarbonate resin.
A visual examination of the chambers confirmed severe cracking within the outlet nipple (Figure 1). The fracture surface was relatively smooth, suggesting slow crack initiation and extension. No signs of apparent ductility were evident. The cracking ran through an area of the part having a greater local wall thickness.
Further inspection of the fracture surface via scanning electron microscopy (SEM) revealed generally brittle fracture features (Figure 2). The examination also indicated that the cracking had initiated along the inner diameter of the nipple wall. Multiple individual crack origins were present. The origin locations exhibited a very smooth morphology, evident of slow crack development.
Analysis of the chamber material, including Fourier transform infrared spectroscopy (FTIR) and differential scanning calorimetry (DSC), produced results characteristic of a polycarbonate resin, with no signs of contamination or degradation. Further, evaluation of the molecular weight of the polycarbonate resin through melt flow rate testing showed no evidence of degradation associated with the molding process. However, thermomechanical analysis (TMA) performed on material adjacent to the fracture surface indicated a high level of molded-in residual stress, as indicated by expansion above the glass transition temperature (Tg).
The investigation concluded that the flow regulator chamber failed via brittle fracture associated with a ductile-to-brittle transition as a result of the exposure of the part to stresses well below the yield strength of the material over an extended period of time. This produced a creep rupture mechanism leading to failure. The design of the housing, including non-uniform wall thickness, produced an area of stress concentration, which is thought to have been a factor in the cracking. The source of the stress that caused the failure was determined to be two-fold in nature. Molded-in residual stress within the nipple area was additive with the side loading external stress imparted on the nipple by the mating hose. The combination of these two stresses led to the ultimate creep rupture, accelerated by the intermittent excursions to elevated temperature. The information generated by the failure analysis was used to correct the problem with the failed component, but also to provide valuable information for future parts.
Putting science into the failure investigation resulted in a timely and cost-effective solution to the flow regulator chamber failure. Time and money was not wasted chasing unproductive theories. In addition, insights were gained which hopefully will lead to avoiding similar problems with future parts.
Published in R & D magazine: Vol. 52, No. 1, February, 2010, pp.16-17.
