The severity of the failure mode is taken into consideration as well as the effect the failure mode, should it occur, will have on the component, system, process, vehicle, or customer. The two main types of FMEAs used in the automotive industry are design FMEAs and process FMEAs.
FMEAs were first introduced in the 1940s in the US military. However, the tool came into its full use when it was used in the 1960s in connection with manned-space missions. The auto industry did not widely adopt FMEAs until the late 1970s. Ford Motor Company introduced them for safety and regulatory items to improve automotive designs and manufacturing, specifically in response to the Pinto fuel tank issues. Today, FMEAs are now starting to make their way into the service side of the business as well. The tool is widely used throughout the aeronautical, automotive, software, food services, health care, and many other industries.
An FMEA is started by listing all of the functions of the item that is being considered. Then the failure modes for each function are listed. A failure mode is an error or defect with the item. For example, if one of the functions of the system does not operate, this is considered a failure mode. If the function only operates some of the type or if an unwanted function occurs instead of the desired function, these are also considered failure modes.
Next, the effect of the failure mode is listed and analyzed. An effect could be minor or it could be major. For example, one major effect of an exhaust system failure could be that the vehicle no longer meets the emissions requirements, and it cannot be sold to customers. Every effect is given a severity rating so that the level of risk can be understood.
Next, the potential causes of the failure mode are listed. One failure mode can have multiple causes. They should all be listed. Each cause is then given an occurrence rating. The occurrence rating tries to give the users of the FMEA an idea of how often the cause can happen.
The FMEA now moves into understanding the control methods for detecting the cause. For example, a design FMEA would list the testing methods that are used to make sure that a part is durable for the life of the vehicle as a current control. The controls are then given a detection rating for how good they are at finding the fault with the system.
Finally, a criticality number and a risk priority number are given to each line in the FMEA. The criticality number is the severity rating multiplied by the occurrence rating. The risk priority number (RPN) is the criticality number multiplied by the detection rating.
To understand the process, here is an example for one line of a brake system design FMEA.
- Function = Vehicle stops when the customer pushes the brake pedal.
- Failure Mode = Vehicle does not stop in the specified distance.
- Effect = Vehicle is unsafe and does not meet the regulatory requirements for stopping distance.
- Severity = 10
- Cause = The brake pad area is not large enough for the weight of the vehicle.
- Occurrence = 2
- Current Controls = Design Guidelines, Stopping Distance Testing, Weight Calculations
- Detection = 1
- Criticality = 20
- RPN = 20
This is just one example for the cause of this particular failure mode. There could be multiple causes and multiple effects for that matter. Each one should be investigated in the FMEA.
Automotive Systems That Use FMEAs in the Design Process
Once every failure mode and cause is identified. The criticality and RPN are then used to prioritize items to investigate further and to develop recommended actions. When FMEAs were originally developed, the RPN number was only used to prioritize items. However, it is generally accepted today that the criticality number should be used first, then severity, then RPN. To help understand the reason for this change, consider the following example:
Effect 1 has a severity of 5, an occurrence of 10, and a detection rate of 2.
Effect 2 has a severity of 2, an occurrence of 5, and a detection rate of 10.
Both of these effects have RPNs of 100. However, effect 1 has a criticality of 50 and effect 2 has a criticality of 10. Although both effects are important to work on, effect 1 has the greater risk in terms of severity and occurrence and should be prioritized over effect 2.
The ultimate purpose of the FMEA is to help designers to eliminate or reduce failures. Because of this goal, the FMEA is never completed. It is a living document that should be constantly used throughout all of the phases of the product life cycle.
As with any program management tool, the FMEA is only has good as the information that is contained it in. To ensure that the FMEA is a high quality document, the FMEA should be completed by a team of people that includes all of the experts involved in the design or process. This approach will allow for a fresh eyes approach and a thorough brainstorming session to be used.
As already discussed above, an FMEA can be used to help improve the quality and reliability of a product or process. However, as with other tools, there are limitations with FMEAs. For example, the FMEA takes a very long time to initially complete and is resource intensive. The failure and effect columns can also be difficult to fill in if some general guidelines are not followed for the entire document. Teams can spend time arguing about the wording and not actually investigating recommended actions. Finally, teams can waste time arguing over the ratings columns. Because of the resource levels required to generate a quality FMEA document, teams often do not revisit them at regular intervals. They are often filed and not used throughout the design process.
Despite these limitations, FMEAs are a valuable tool and can help designers and engineers improve products and services within the automotive industry.