Failure Mode and Effect Analysis Enhances Reliability by Detecting Problems Early
By Erik FadlovichErik Fadlovich is a quality control expert with more than 13 years experience in the manufacturing industry. Fadlovich holds a bachelor's degree in Aviation Technologies and Operations from Western Michigan University. His current duties as Watlow's Quality Manger include profit and loss management, continuous operations improvement, project management and training for the ISO9001:2000 and TS16949:2002 quality system.
Failure Mode and Effect Analysis (FMEA) is methodology for analyzing potential problems early in the product development cycle where it is easier to take action to overcome potential issues, thereby enhancing reliability through design. The tool is used to identify relationships between process and product requirements, as well as the potential for unacceptable outputs and their effects.
FMEA takes time to perform and, if done accurately, provides a valuable tool to plan, detect, and react to ensure a successful product lifecycle. Continuing to use the tool throughout the product lifecycle will significantly improve safety, quality, delivery, and cost. Additional benefits from the tool can be gained through cause chain analysis and mistake proofing (i.e., poke yoke) to reduce Risk Priority Numbers (RPN), which is discussed later in this article.
The FMEA is developed prior to the launch of a new job or process and is maintained throughout its life. As a tool embedded in Six Sigma methodologies, FMEA also helps identify the controls (through development or production) that must be implemented to ensure that the product can be produced continuously within specification. During the life of a product/process, the tool (considered a living document) should be modified whenever an existing design, product, or process is changed, or even when a derivative is in development. Furthermore, the tools should be used when a new design, product, or process is in the discovery stages.
DFMEA and PFMEA
There are several applications for FMEA, including: design, which focuses on components and subsystems; process, for manufacturing and assembly processes; system, which orients on global system functions; service functions; and software functions. This article will focus on two primary formats: DFMEA (design) and PFMEA (process). These documents both help to manage the risks for customers.
Once the DFMEA is completed, the PFMEA can be developed to control the production process. At Watlow, PFMEA documents are developed based on their part numbers. This means that the PFMEA must list the steps for the building of a part from receipt of the product through to shipping. The disadvantage of this approach is that when a change is made to a process that is common to more than one part number, all PFMEA documents containing references to that process must also be changed. Using a process structure (versus the part number structure), it becomes clear as to the benefits gained by focusing on the process. Take Watlow's induction braze process as an example. If a PFMEA is developed for induction brazing, it can be applied to all part numbers that utilize the induction braze process. When improvements are made to the induction braze process, updates are made to the PFMEA specific to this process. The benefit is that only one change must be made on one document to cover all part numbers that use this process. Additionally, the PFMEA can be used to prioritize improvement using an RPN for a process that relates to many part numbers.
Risk Priority Number
Using FMEA, an RPN can be derived by determining the severity of the potential failure mode, the possibility of occurrence, and the likelihood that a defect will or will not be detected. Once the FMEA is understood, it is easy to determine which area has the greatest concern, and therefore, take action to prevent the problem before it arises.It is also important to know how the RPN is calculated. In the example below, the process function/requirement that has been identified is the joining of similar metals via solder process. Moving to the right, the potential failure mode is incomplete braze joint. The effect that each potential failure mode may have on the product is listed in the potential effect(s) of failure column (intermittent function and inoperable function), and is ranked with a severity (SEV) number. The severity number is ranked one to ten with ten being the most severe and one having no effect.
The potential cause(s)/mechanism(s) of failure can the be assessed. For example, improper setup (e.g., power setting, coil size, time setting, etc.), improper coil condition, and improper flux application (Figure 2) to determine the failure occurrence (OCCUR). This number is also ranked from one to ten with ten having an almost inevitable chance of failure and one having a very remote chance of failure. The current process control can then be listed and the possibility of detection (DETEC) that these controls will prevent can be detected. Each item is ranked one to ten, however, the ranking is reversed where ten is almost impossible to detect and one almost certain. This is then repeated down the PFMEA form.
After ranking the severity, occurrence, and detection modes, the risk can be calculated by multiplying the three numbers and placing the result under the RPN column. Once this is done for the entire design and/or process, the areas of greatest concern are readily identified by simply comparing the RPN numbers. The process(es) that have the highest RPN should be given the highest priority for corrective action. In the example in Figure 2, the greatest area of concern would be documented work instructions to prevent intermittent function as a result of an improper set up (RPN = 512).
Often, it is described to address situations which are the most severe (inoperable function with a SEV ranking of 9). However, this may not always be true. For instance, the intermittent function (SEV ranking of 8) is identified as having the greatest risk (RPN of 512) because it is more likely to occur (OCCUR ranking of 8) and less detectable (DETEC ranking of 8). In conjunction, it is critical to ensure that the PFMEA focus is on process controls (e.g., poke yoke, statistical control of critical equipment parameters, etc.) that prevent potential failure modes.
To complete the PFMEA, recommended actions with targets and dates of implementation are noted. Once the actions have been implemented, the severity, occurrence, and detection are again ranked in the far right of the form and actions prioritized accordingly. These numbers can easily be put into a graph, providing a visual representation of activities that should be prioritized to mitigate risk.The use of the DFMEA to effectively develop a product design from concept to production further allows for an effective PFMEA to be created and for processes to be deployed. Using the two documents interactively (remember, they are living documents and will be revised throughout the life of the product), has helped Watlow meet customer specifications while reducing design changes and scrap, improving delivery, and controlling costs.