MACHINERY FMEA (MFMEA) | Six Sigma and Beyond: Design for Six Sigma, Volume VI

A machinery FMEA is a systematic approach that applies the traditional tabular method to aid the thought process used by simultaneous engineering teams to identify the machine's potential failure modes, potential effects, and potential causes of the potential failure modes and to develop corrective action plans that will remove or reduce the impact of the potential failure modes. Generally , the delivery of a MFMEA is the responsibility of the supplier who generates a functional MFMEA for system and subsystem levels. This is in contrast to a DFMEA where the responsibility is still on the supplier but now the focus is to generate transfer mechanisms, spindles, switches, cylinders , exclusive of assembly-level equipment.

A typical MFMEA follows a hierarchical model in that it divides the machine into subsystems, assemblies, and lowest replaceable units. For example:

Level 1: System level ” generic machine
Level 2: Subsystem level ” electrical, mechanical, controls
Level 3: Assembly level ” fixtures/tools, material handling, drives
Level 4: Component level
And so on

IDENTIFY THE SCOPE OF THE MFMEA

Use the boundary diagram. Once the diagram has been completed, you can focus the MFMEA on the low MTBF and reliability values.

IDENTIFY THE FUNCTION

Define the function in terms of an active verb and a noun. Use a functional diagram or the P diagram to find the ideal function. Always focus on the intent of the system, subsystem, or component under investigation.

FAILURE MODE

A failure is an event when the equipment/machinery is not capable of producing parts at specific conditions when scheduled or is not capable of producing parts or performing scheduled operations to specifications. Machinery failure modes can occur in three ways:

Component defect (hard failure)
Failure observation (potential)
Abnormality of performance ( constitutes the equipment as failed)

POTENTIAL EFFECTS

The consequence of a failure mode on the subsystem is described in terms of safety and the big seven losses. (The big seven losses may be identified through warranty or historical data.)

Describe the potential effects in terms of downtime, scrap, and safety issues. If a functional approach is used, then list the causes first before developing the effects listing. Associated with the potential effects is the severity, which is a rating corresponding to the seriousness of the effect of a potential machinery failure mode. Typical descriptions are:

Downtime

Breakdowns: Losses that are a result of a functional loss or function reduction on a piece of machine requiring maintenance intervention.
Setup and adjustment: Losses that are a result of set procedures. Adjustments include the amount of time production is stopped to adjust process or machine to avoid defect and yield losses, requiring operator or job setter intervention.
Startup losses: Losses that occur during the early stages of production after extended shutdowns (weekends, holidays, or between shifts), resulting in decreased yield or increased scrap and defects.
Idling and minor stoppage: Losses that are a result of minor interruptions in the process flow, such as a process part jammed in a chute or a limit switch sticking , etc., requiring only operator or job setter intervention. Idling is a result of process flow blockage (downstream of the focus operation) or starvation (upstream of the focus operation). Idling can only be resolved by looking at the entire line/system.
Reduced cycle: Losses that are a result of differences between the ideal cycle time of a piece of machinery and its actual cycle time.

Scrap

Defective parts: Losses that are a result of process part quality defects resulting in rework , repair, or scrap.
Tooling: Losses that are a result of tooling failures/ breakage or deterioration/wear (e.g., cutting tools, fixtures, welding tips, punches, etc.).

Safety

Safety considerations: Immediate life or limb threatening hazard or minor hazard .

SEVERITY RATING

Severity is comprised of three components :

Safety of the machinery operator (primary concern)
Product scrap
Machinery downtime

A rating should be established for each effect listed. Rate the most serious effect. Begin the analysis with the function of the subsystem that will affect safety, government regulations, and downtime of the equipment. A very important point here is the fact that a reduction in severity rating may be accomplished only through a design change. A typical rating is shown in Table 6.9.

Table 6.9: Machinery Guidelines for Severity, Occurrence, and Detection
Effect	Criteria Severity	Rank	Probability of Failure	Criteria for Occurrence	Rank	Alternate Criteria for Occurrence	Detection	Criteria for Detection	Rank
Hazardous without warning	Very high severity: affects operator, plant, or maintenance personnel safety and/or effects noncompliance with government regulations without warning	10	Failure occurs every hour	R(t) < 1 or some MTBF	10	1 in 1	Very low	Present design controls cannot detect potential cause or no design control available	10
Hazardous with warning	High severity: affects operator, plant or maintenance personnel safety and/or effects noncompliance with government regulations with warning	9	Failure occurs every shift	R(t) = 5%	9	1 in 8		Team's discretion depending on machine and situation	9
Very high	Downtime of 8+ hours or the production of defective parts for over 2 hours	8	Failure occurs every day	R(t) = 20%	8	1 in 24		Team's discretion depending on machine and situation	8
High	Downtime of 2 “4 hours or the production of defective parts for up to 2 hours	7	Failure occurs every week	R(t) 37%	7	1 in 80	Low	Machinery control will isolate the cause and failure mode after the failure has occurred, but will not prevent the failure from occurring	7
Moderate	Downtime of 60 “120 min or the production of defective parts for up to 60 min	6	Failure occurs every month	R(t) = 60%	6	1 in 350		Team's discretion depending on machine and situation	6
Low	Downtime of 30 “60 min with no production of defective parts or the production of defective parts for up to 30 min	5	Failure occurs every 3 months	R(t) = 78%	5	1 in 1000	Medium	Machinery controls will provide an indicator of imminent failure	5
Very low	Downtime of 15 “30 min with no production of defective parts	4	Failure occurs every 6 months	R(t) = 85%	4	1 in 2500		Team's discretion depending on machine and situation	4
Minor	Downtime up to 15 min with no production of defective parts	3	Failure occurs every year	R(t) = 90%	3	1 in 5000	High	Machinery controls will prevent an imminent failure and isolate the cause	3
Very minor	Process parameter variability not within specification limits. Adjustments may be done during production; no downtime and no defects produced	2	Failure occurs every 2 years	R(t) = 95%	2	1 in 10,000		Team's discretion depending on machine and situation	2
None	Process parameter variability within specification limits; adjustments may be performed during normal maintenance	1	Failure occurs every 5 years	R(t) = 98%	1	1 in 25,000	Very high	Machinery controls not required; design controls will detect a potential cause and subsequent failure almost every time	1

It should be noted that these guidelines may be modified to reflect specific situations. Also, the basis for the criteria may be changed to reflect the specificity of the machine and its real world usage.

CLASSIFICATION

The classification column is not typically used in the MFMEA process but should be addressed if related to safety or noncompliance with government regulations. Address the failure modes with a severity rating of 9 or 10. Failure modes that affect worker safety will require a design change. Enter "OS" in the class column. OS (operator safety) means that this potential effect of failure is critical and needs to be addressed by the equipment supplier. Other notations can be used but should be approved by the equipment user .

POTENTIAL CAUSES

The potential causes should be identified as design deficiencies. These could translate as:

Design variations, design margins, environmental, or defective components
Variation during the build/install phases of the equipment that can be corrected or controlled

Identify first level causes that will cause the failure mode. Data for the development of the potential causes of failure can be obtained from:

Surrogate MFMEA
Failure logs
Interface matrix (focusing on physical proximity, energy transfer, material, information transfer)
Warranty data
Concern reports (things gone wrong, things gone right)
Test reports
Field service reports

OCCURRENCE RATINGS

Occurrence is the rating corresponding to the likelihood of the failure mode occurring within a certain period of time ” see Table 6.8. The following should be considered when developing the occurrence ratings:

Each cause listed requires an occurrence rating.
Controls can be used that will prevent or minimize the likelihood that the failure cause will occur but should not be used to estimate the occurrence rating.

Data to establish the occurrence ratings should be obtained from:

Service data
MTBF data
Failure logs
Maintenance records

SURROGATE MFMEAs

Current Controls

Current controls are described as being those items that will be able to detect the failure mode or the causes of failure. Controls can be either design controls or process controls.

A design control is based on tests or other mechanisms used during the design stage to detect failures. Process controls are those used to alert the plant personnel that a failure has occurred. Current controls are generally described as devices to:

Prevent the cause/mechanism failure mode from occurring
Reduce the rate of occurrence of the failure mode
Detect the failure mode
Detect the failure mode and implement corrective design action

Detection Rating

Detection rating is the method used to rate the effectiveness of the control to detect the potential failure mode or cause. The scale for ranking these methods is based on a 1 to 10 scale ” see Table 6.8.

RISK PRIORITY NUMBER (RPN)

The RPN is a method used by the MFMEA team to rank the various failure modes of the equipment. This ranking allows the team to attack the highest probability of failure and remove it before the equipment leaves the supplier floor.

The RPN typically:

Has no value or meaning (Ratings and RPNs in themselves have no value or meaning. They should be used only to prioritize the machine's potential design weakness [failure mode] for consideration of possible design actions to eliminate the failures or make them maintainable .)
Is used to prioritize potential design weaknesses (root causes) for consideration of possible design actions
Is the product of severity, occurrence and detection (RPN = S — O — D)

Special note on risk identification:

Whereas it is true that most organizations using FMEA guidelines use the RPN for identifying the risk priority, some do not follow that path. Instead, they use a three path approach based on:

Step 1: severity
Step 2: criticality
Step 3: detection

This means that regardless of the RPN, the priority is based on the highest severity first, especially if it is a 9 or a 10, followed by the criticality, which is the product of severity and occurrence, and then the RPN.

RECOMMENDED ACTIONS

Each RPN value should have a recommended action listed.
The actions are designed to reduce severity, occurrence, and detection ratings.
Actions should address in order the following concerns:
Failure modes with a severity of 9 or 10
Failure mode/cause that has a high severity occurrence rating
Failure mode/cause/design control that has a high RPN rating
When a failure mode/cause has a severity rating of 9 or 10, the design action must be considered before the engineering release to eliminate safety concerns.

DATE, RESPONSIBLE PARTY

Document the person, department, and date for completion of the recommended action.
Always place the responsible party's name in this area.

ACTIONS TAKEN/REVISED RPN

After each action has been taken, document the action.
Results of an effective MFMEA will reduce or eliminate equipment downtime.
The supplier is responsible for updating the MFMEA. The MFMEA is a living document. It should reflect the latest design level and latest design actions.
Any equipment design changes need to be communicated to the MFMEA team.

REVISED RPN

Recalculate S, O, and D after the action taken has been completed. Always remember that only a change in design can change the severity. Occurrence may be changed by a design change or a redundant system. Detection may be changed by a design change or better testing or better design control.
MFMEA ” A team needs to review the new RPN and determine if additional design actions are necessary.