A FMEA is a stable and seasoned design tool, the Failure Mode Effects Analysis (FMEA) Study. Often called a Failure Mode Study, this design and maintenance engineering tool has existed and matured for some time, but is still ignored by many design teams and organizations.

While the last thing design, development and maintenance engineering needs is another acronym, Failure Mode and Effects Analysis (FMEA) may be one of the easiest concepts to understand and apply. Best of all, it delivers significant financial returns quickly.

Failure Mode & Effects Analysis is an approach that helps identify and prioritize potential equipment and process failures. Often used as a springboard to establishing a Root Cause Failure Analysis program, FMEA is a logical system that objectively ranks potential failures and provides recommendations for corrective actions.

Is This Really Necessary ?

Informal failure analysis happens every day in most plants as engineers try to figure out why a machine unexpectedly broke down or how a part slowly came out of tolerance.

By contrast, FMEA is a formal process that allows in-house experts to concentrate on failures and fix them.

Failure Mode and Effects Analysis is important because it focuses on failures and potential problems. If it's done early enough in the process, we can anticipate problems and engineer them out of the system. And the earlier you catch potential failures, the more money you save.

3 Types of FMEAs

1) Design – Design FMEAs strive to eliminate causes of failure during equipment design, taking into account everything from ease of maintenance to potential safety concerns.

2) Process – Process FMEAs focus on problems stemming from how the equipment is maintained and operated.

3) System – System FMEAs look for potential failures and bottlenecks in larger processes, such as entire production lines.

Many times the three levels overlap. They often happen simultaneously, particularly design and process FMEAs, because so much of the design depends on how the equipment is handled, and how it gets handled depends on design.

Maintenance engineers are most often involved in process FMEAs, because equipment is usually installed and running by the time maintenance is asked to determine why a failure occurred.

In a process FMEA, maintenance's equipment expertise is augmented by experts in other plant functions. A maintenance-oriented process FMEA might be conducted by a team consisting of a process engineer, a machinery operator and two maintenance people.

A cross-functional team of at least four members is recommended for any type of FMEA.

Cross-functional teams are necessary for two reasons:

1) No plant function operates in a vacuum and it makes sense to get perspectives from other team members.

2) Specific application expertise.  Education, background, work experience, familiarity with past systems and failures.  Each team member brings in another level of expertise.

Another reason for teams lies at the heart of the FMEA process: ratings. FMEAs are, essentially, a method of rating failure occurrances and severity and the chances of detecting and correcting the failures before they cause harm.

How a FMEA Works

1) Team isolates and describes the potential failure modes.

2) Team discusses the potential effect of each failure.

3) Team must assess the severity, occurrence and detection of failures and give those aspects different and meaningful numeric ratings. Ratings are typically from 1 to 10, with 1 being the least severe, least occurring, or most easily detectable. 10 would be those faults that are most severe, most catastrophic and those hardest to detect.

Isolate and Describe

To determine a particular piece of equipment's risk of failure, team members must first isolate and describe the potential failure mode: under what conditions does the equipment fail?

In the case of a maintenance-related problem, this is often as simple as reviewing equipment history.

In a design FMEA, this could include failure modes in the hardware, software, mechanical and system portions of the design.

Discuss Potential Effect

Next, the team discusses the potential effect of that failure: Does normal use of the product stop? Is production halted? Is product quality affected? Does someone get hurt?

Create Severity, Occurrence, and Detection Ratings

Once the failure mode has been defined and the potential effects of failure have been determined, the team must assess the severity, occurrence and detection of failures and give those aspects numeric ratings.

Severity means how serious the failure will be. You give it a rating from one to ten, where ten is the most severe failure.

In the same way, you rate the occurrence of failure - how frequently you see the failure.
 
Detection indicates how easily that fault or failure can be detected. The detection scale is the reverse of the other 2 scales, with 1 being the easiest or most detected and 10 being the hardest or most difficult to detect.
 
Obviously, for this rating system to work, it's vital that all team members understand what constitutes a failure. Each potential effect of failure is given a severity, occurrence and detection rating. Those numbers are multiplied to produce a Risk Priority Number (RPN).
 
Example: One potential failure is that a worker gets his arm cut off in a process - that's a very high severity rating, a 10. The occurrence of that is very low, however, and is given a 1. Detection is obvious, so that also gets a 1. So the RPN is 10 (10 times 1 times 1).
 
Premature die wear is another example of a failure mode. It's a fair severity, but we can catch it before it affects the customer - say it's a 6. But the frequency of seeing that is much higher, maybe also a 6. The ability to detect this type of wear may vary but we assume we have a good maintenance process in place and we give this a 3. So the RPN for premature die wear is 6 times 6 times 3.

At 108, the problem of premature die wear is more likely to happen than the problem of losing an arm, with an RPN of 10. So the die wear problem should be addressed first.
 
The RPN is an absolute ranking of a system's potential for failure. After several FMEAs are undertaken, the RPNs reveal a clear pecking order for prioritizing corrective actions. The RPN number also serves as a benchmark against which to measure improvements.

Prioritizing, Actions Plans

Once the RPN is calculated, the team must develop action plans for correcting, mitigating or eliminating the potential problem.
 
Going through the process of defining possible failure modes and causes of failure gives the team a significant head start on recommending action plans. Usually it's just a matter of translating what they've found into specific tasks and assigning responsibility for getting the job done.
 
When it comes to corrective actions, there might be three or four solutions, but the company might only have the time and money to complete one. So they have to come back when they have the resources to complete the job. In this way, FMEA becomes a living document - it's something you work on over time to complete.

There's no doubt that the effort is worth the time and trouble, however. By focusing on potential failures and their causes, design, maintenance and plant engineers can anticipate and eliminate costly problems before they happen. The savings are enormous.

We can help with your FMEA study.  Go here for details.