Good FMEA / Bad FMEA William... · · 2016-06-18Good FMEA / Bad FMEA By: William Mulligan Session 23E. ... the Apollo missions. 1970s • MIL-STD-1629 ... Material softens with

October 22, 2014 ASQ NEQC 60th Conference, Springfield, Massachusetts

Good FMEA / Bad FMEA

By:

William Mulligan

Session 23E

January 28th, 1986

74.201

Seconds76 +

Seconds

October 22, 2014 ASQ NEQC 60th Conference, Springfield, Massachusetts 2

The Overall Purpose of FMEA is to Lower Risk

3

D. H. Stamatis, FMEA:FMEA from Theory

to Practice, Quality Press, 1995

Unclear Customer

Expectations (AFMEA)

Potential Safety

Hazards (ALL)

Insufficient Knowledge of

Environmental Requirements

(ALL)

Poor Design

(DFMEA)

Raw Material

Variation

(PFMEA)

Poorly developed

Specification Limits

(DFMEA / AFMEA)Measurement

Variation (PFMEA)Machine Reliability

(PFMEA)

Poor Process Control

(PFMEA)

Cumulative Risk

Risk can derive from:


History

1990s - TodayISO 14971: Risk Management for Medical & Diagnostic Devices.

4

1960s:

First used in the

Aerospace industry during

the Apollo missions.

1970s• MIL-STD-1629

• Automotive applications

driven by liability costs

1980sAutomotive industry used FMEA’s to

improve quality and product

reliability.


Definition - FMEA

A bottom-up structured approach to:

– Identify the ways in which a product can fail. Specifically, the Root Cause(s) of failure, the Failure Mode(s), and the Effect(s) of failure.

–Estimate the risk associated with each Failure Scenario.

–Rank-order the Failure Scenarios to prioritize the design team’s efforts.

–Track corrective actions and provide a permanent record for subsequent FMEAs.

The FMEA is a living document which is updated throughout the life of the product.

5

Primary Directive: Identify ways that a product can fail and take

action to remove the failure modes before they occur.


Types of FMEA

Process (PFMEA): Analyzes errors in manufacturing and assembly

processes:

Focuses on process inputs (what can go wrong in the manufacturing process).

Defective components are a consideration.

Design (DFMEA): Analyzes the product design before release to production:

Assumes that the product is properly manufactured to design specifications.

Focuses on product function – suitability of components, geometry, materials, etc. to

produce the desired function.

Application (AFMEA): Analyzes the suitability of a product to meet the

needs and expectations of the customer:

Anticipated and “off-label” uses are considered as well as shipping, storage, disposal.

Assumes that the product is properly manufactured to design specifications.


7

Incorrect or

Defective Design

(DFMEA)

Process Defect

(PFMEA)

Product cannot

meet performance

targets –

anitcipated use

Product is difficult

to manufacture

Inappropriate

Design (AFMEA)

Product is

awkward to use or

non-intuitive

Product is

dangerous if

accidentally

misused

Performance

degrades over

time

Unreasonable

tolerances or

material properties

Product appearance or

performance degrades

due to exposure to

unanticipated

environmental

conditions

Unclear instructions

for use, or

inadequate warnings

Product dangerous

if dropped /

impacted / tipped

over etc.

Performance

varies from

product to product

Results in

Which means

Results in

Non-robust

manufacturing

techniques

required

Which means

Results in

Ineffective, Inappropriate, or Incorrect

Design Inputs

Which can lead to

Which can lead to

A process defect

may cause any of

the results above

Contributing factor

Product

appearance or

performance

inadequate during

unanticipated use.

link

Performance

varies under

anticipated

conditions

Results in

Results in

Which means

Functionality

issuesIncorrect assembly

Missing parts or

accessories

Dirt / marks /

particulate matter

Caused by

Inappropriate mfg

environment

Out-of-spec raw

materials or

components

Process not

validated, fixtures

not poka-yoked,

training issues,

design issues

Inadequate

training, design

issues

How all three FMEA types relate to each other


FMEA Inputs and Outputs

•Inputs:

Product concept design

QFD

Experience with related products

Benchmarking

Product performance requirements

•Outputs:

Prioritized list of actions to prevent / minimize Root Causes or to detect Failure Modes

History of actions taken

8October 22, 2014 ASQ NEQC 60th Conference, Springfield, Massachusetts

Conducting a DFMEA: Part 1

1. Framework: Develop a block diagram of the Product / Component / Function being analyzed. For complex products, a breakdown to sub-system may be adequate.

2. Populate: Enter the Components and functions into the DFMEA template. Each component (or subsystem) function will be analyzed as a separate line item.

3. Brainstorm: Identify the “Chains-of-Events” that may occur as a result of design issues. Think in terms of Root Causes driving Failure Modes driving Effects.

– Root Cause: The design-related source of the issue.

– Failure Mode: What physically happens to the component / subsystem / product.

– Effect: What the user or patient perceives as a result of the failure.

Step 1 Step 2 Step 3


Pencil

Lead Wood Body Eraser Eraser Retainer Paint

Produces visible

marks

Cuts without

cracking

Absorbs

mechnical stress

Retains lead

Cuts without

cracking

Absorbs

mechnical stress

Removes lead

markings

Retains eraser

to body

Provides bright

finish

Provides

continuous

coating for body

Cosmetic

features

10

Every failure is associated with a Function. (What is the product not doing that it is supposed

to do?) The Root Cause / Failure Mode / Effects are each associated with a specific Product

or Component function.

When creating the DFMEA, each component or subsystem is listed, and each function of the

component or subsystem is analyzed separately for “what could go wrong” with that function.

One product per DFMEA

DFMEA sections are

defined by Components

Line Items are defined by

component function.

Each function of each

component is analyzed

separately.


Definition of Terms – Root Causes, Failure Modes, Effects:

11

It is important that the difference between Root Cause, Failure Mode, and

Effect of Failure is clearly understood. These form the “Chain of Events”:

• Root Cause: The source of design errors, or a design weakness that

causes the Failure Mode to occur.

• Failure Mode: The way in which a specific product feature fails to meet

the designed intent. If the Failure Mode is not detected and either

corrected or removed, it will cause the Effect of Failure to occur. The

Failure Mode can also be the Root Cause of a potential Failure Mode in

a higher level assembly.

• Effect: The impact on the item’s functionality and performance to

designed intent as perceived by the customer.

Remember: A customer is not only the end user: Customers may include

Manufacturing, higher level assembly design groups, Field Service.


Root Causes, Failure Modes, and Effects of Failure: Challenger Space Shuttle, January 28, 1986

12

Failure Mode

Escape of hot exhaust

gasses through o-ring

Root Cause

Inappropriate o-rings for

low ambient temperature

operation. Poor sealing.

Effect

Ignition of fuel tank

Function: O-Rings seal exhaust gases.


Chain of Events Thinking:

13

There may be more than one Root Cause for a given Failure Mode, more than one

Failure Mode for a given Effect, and more than one Effect for a given Failure Mode:

POTENTIAL FAILURE

Root Cause... Failure Mode...

...Effect to user

...Effect to patient

drivesresults in

results in

(may also be driven by other Root Causes)

(may also cause other Effects)

(may also be the result of

other Failure Modes)

(may also be the result of

other Failure Modes)(may also drive other Failure Modes)

Surgical product example


Chain of Events Thinking and the “Window of Consideration”

14

Best choice for Window of Consideration

Speculative SpeculativeLikely EffectTarget of C/A Root Cause Failure Mode

Note that the “Patient injury” and “Legal Action” results are speculative and depend

on factors outside the chain of failure. Likewise with the training of the designer. The

appropriate window, providing the most useful information is shown in the blue box.

Insufficient material

finish spec’d for Pin

Excessive friction,

pawl binds on pin

during use

Instrument remains

locked on tissue

Inadequate training

of designer on

materials or

mechanism

Patient injury

Legal action against

hospital or

company


Root Cause: Pitfalls and Pointers

15

Material softens with use Material deforms under load

Too much clearance Potential mechanical interference

Electrical overstress Rough surface finish

Leads are incorrect length for application

Material yellows with age

Material too brittle Specified lubricant is inadequate for product life

Pitfalls

Do not “blame” the environment:

NOT: The dishwasher caused the stress crack at the interface.

RATHER: The material chosen degrades in the dishwasher.

Avoid vague responses, such as “incorrect geometry”, or “improper tolerancing”.

Do not include Root Causes due to manufacturing errors or defective components—assume product is built to the design, but design may not be optimized.

Potential Root Cause:Material Chosen inadequate for dishwasher

environment (becomes brittle)

Dishwasher

Some Examples:

Pointers

Keep in mind the Usage Environment & Reasonable Abuse


Failure Mode: Pitfalls and Pointers

16

Function: Support the cup when

lifted by the user

Failure Mode:Handle cracks / fractures at

cup interface

Element: Handle

Pitfall:Confusing the Failure Mode with Root Cause or Effect.

For Root Cause, ask yourself ‘What is wrong with the design that might cause a problem?’ The answers usually include words that involve geometry and/or materials. Examples are; “inadequate strength”, “inadequate clearance/interference”, “too large”, “too soft”, “unsuitable for environment”.

For Failure Mode, ask yourself ‘What happens to the instrument itself as a result of the design issue?’. The answers usually include words like; “fractures”, “loosens”, “corrodes”, “deforms”, “detaches”, “jams”.

Some Examples:

Delamination Deformed

Unintentional Input Cracked

Fractured Loosened

JammedOxidized

Open Circuit

Short to Ground

Pointer:Having very clear and concise functions really helps thebrainstorming process because it naturally lends itself toFailure Modes and Root Causes.


Effect: Pitfalls and Pointers

17

Effect: “Cup detaches

from handle when lifted by user.”

Pitfalls

Do not address harm: NOT: Burned User (We do not know if the user will, in fact, be burned.)

Do not be speculative: NOT: Hot Coffee Spilled (How do we know that the coffee cup was filled with a hot liquid when the handle detached?)

Some Examples:

Fluid or air leakExposed sharp edge

Knife can not be deployedPoor staple formation

Dosing ErrorParts detach during use

Cannot Adjust Power LevelUser-perceived noise or vibration

Particulate in Field

Rattle

Pointers: Describe the potential Effects of the Failure Mode in sufficient detail such that

Severity can be determined.

There may be multiple Effects and Root Causes that need to be considered.


Applied Chain of Events Thinking

Root Cause Failure Mode Effect

Incompatible raw materials –

mating parts

Galvanic corrosion Parts jammed together

Inadequate lubrication type or

quantity specified

Excessive Wear Excessive play in mechanism –

early failure

Inappropriate fastener material

specified (be specific)

Parts loosen during use Rattle

Insufficient clearance in

tolerance stack up

Interference between parts Potential difficulty assembling

product, malfunction during use.

Inappropriate strength material

specified (too weak)

Material yields under load Mechanical failure of system

Inappropriate wire gauge

specified (too thin)

Wiring temperature excessive

during extended use

Potential distortion / ignition of

surrounding components

Component tolerancing

unrealistic (too tight)

Non-robust manufacturing

techniques required

Assembly and Fabrication

issues (drives PFMEA)

18

Generic examples of Root Causes, Failure Modes and associated Effects:

Note that these Root Cause examples are generic…yours should be specific:

Avoid the terms “improper”, “incorrect”, “inappropriate”, “inadequate” and the

like in your actual DFMEA.


Function Root Cause Failure Mode Effect

Length(s) of

tubing

Incorrect lengths

defined by design

input.

Incorrect lengths Lengths not correct for application

Application

Step


Surgeon

places

instrument

through tube

Tubing length too

long

Instrument does not

protrude sufficiently from

tubing for application

Inadequate maneuverability of

instrument

19

This is more of an AFMEA and Voice-of-Customer issue. Why did we design something

to an incorrect size? The description of the Component Function is also vague. Further,

the Failure Mode and Effect do not add any value. If the item was moved to AFMEA, it

might look like this:


Component

Function


Seal – must

resist

repeated inst

insertions

Improper design Seal torn / damage Air leakage, Fluid Leakage, Pump

can’t maintain pressure

Component

Function


Seal – resists

repeated inst

insertions

Seal material too

brittle for application

Seal tears Air Leakage

(list other Effects separately)

20

In this case, the Root Cause is way too vague – what about the design of the seal

causes failure? Is it the material, geometry, assembly? The Failure Mode is OK, but

there are many Effects listed which may have differing severities.


Component

Function


Knife – cutting

ability

Burr blade Insufficient knife edge Rough or jagged edge cut

Component

Function


Knife – cutting

ability

Blade specification

does not preclude

burrs

Uncontrolled knife edge,

burrs present

Rough or jagged edge cut

21

This appears to be a PFMEA item – there is an apparent manufacturing defect in the

blade. Further, the Failure Mode doesn’t add anything. The Effect is OK, describing the

result in terms of the user. If the issue is that the edge condition is not spec’d properly,

this may be the entry:


Component

Function


Blunt edge of

instrument to

prevent tissue

trauma

Cracks, sharp

geometry

Sharp Hazard Patient Injury

Component

Function


Blunt edge of

instrument is

non-traumatic


strength specified

Edge cracks during

application to tissue

Sharp hazard in patient cavity

22

The Root Cause is really a Failure Mode, the Effect is speculative and non-descriptive

and the Failure Mode is really an Effect. What’s missing is the real Root Cause – the

Window of Consideration should be adjusted:


Component

Function


Pusher bar

feeds clip into

jaw slots

Clip path obstructed Pusher bar buckles Instrument jams

Application

Step


Squeeze

handle to feed

clip

Clip path is

obstructed by

patient anatomy

Pusher bar buckles Instrument jams

23

This is OK except for the Root Cause: What is the source of the obstruction? What is the

design related issue? If the surgeon or patient anatomy is obstructing the clip path, this

becomes an AFMEA item as shown below:


1. List the Current Controls for each Chain of Events.

2. Determine the Severity, Occurrence, and Detection rating scales.

3. Assign Severity, Occurrence and Detection scores to each line item.

4. Calculate the RPN for each line item.

5. Determine Recommended Actions to reduce high RPNs and mitigate risk.

6. Take appropriate actions, record outcomes, and recalculate RPNs.

7. Review AFMEA and PFMEA to determine if additional updates to those documents are required.

Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7


Controls are systematized methods put in place to prevent or detect Failure Modes or Root

Causes before Effects occur. Controls can either prevent the Root Cause or Failure Mode

from occurring, or detect the Failure Mode prior to the Effect occurring.

Root Cause Failure Mode Effect Control

Incompatible raw

materials – mating parts

Galvanic corrosion Mechanical Failure Material Review

Inappropriate lubrication

specified

Excessive Wear Excessive play in

mechanism – early failure

Pre-production life test with

specified lubricant

Inappropriate fastener

material specified

Parts loosen during use Rattle Engineering test of fastener

strength, vibratory loading

Inappropriate tolerance

stackup

Interference between

parts

Potential difficulty

assembling product,

malfunction during use.

Tolerance study of

components based on actual

Ppk data.

Inappropriate strength

material specified

Material yields under

load

Mechanical failure of

system

Material stress / strength

testing

Inappropriate wire

gauge specified (too

thin)

Wiring temperature

excessive during

extended use

Potential fire hazard Life testing with amp and

temperature monitoring

Inappropriate

tolerancing of

component

Non-robust

manufacturing

techniques required

Assembly and Fabrication

issues (drives PFMEA)

Design review with

manufacturing to assess

fabrication difficulty

October 22, 2014 25ASQ NEQC 60th Conference, Springfield, Massachusetts

Example: Pin (component of larger assy)

Jaws remain locked

on vesselInsufficient material


A B C D E F G H I J K

Ite

m #

Component / Function Potential Failure

Mode(s)

Potential Effect(s) of

Failure

Se

ve

rity

Potential Root

Cause(s)

Oc

cu

rre

nc

e

Current Design

Control(s)

Co

ntr

ol

sta

tus

De

tec

tio

n

RP

N

COMPONENT #1

1 1st Function of Component #1 0

2 2nd Function of Component #1 0

3 0

4 0

COMPONENT #2



7 0

8 0

COMPONENT #3



11 0

12 0

COMPONENT #4



15 0

16 0


finish spec’d for PinActuates pawlExcessive friction,

pawl binds on pin

during use

Instrument remains

locked on tissue

Test to determine

functional life of

matl finish

PIN


For the 1-10 scoring scale, the following applies:

Severity (of Effect): Given that the failure occurs, score the importance of the Effect to the customer. The Severity score must include potential safety risks if failure occurs:

1=Not severe: Customer will barely notice failure

10=Very Severe: Effect may be life threatening

Occurrence (of Cause / Failure Mode / Effect): The frequency with which a given Root Cause occurs and creates the Failure Mode, resulting in the Effect:

1=Unlikely to occur

10=Will occur frequently

Detection (effectiveness of Current Design Controls): The ability of the current controls to prevent the Cause(s) or detect the Failure Modes:

1=Highly likely to Prevent / Detect,

10=Not likely at all to Prevent / Detect

Note the relationship between the prevention aspect of the Current Controls and

the Occurrence score. A Current Control which effectively prevents the occurrence

of the Root Cause or Failure Mode will force a low Occurrence score.


Example scoring scale for Occurrence, Detection, Severity:

Rating Severity Occurrence Detection

10

Dangerously High, May injure the

customer, Hazardous without

warning.

Can not detect

9Extremely High, Potentially

hazardous with warning.

Very remote chance of detection

8Very high, Loss of primary

function, renders unfit for use.

Remote chance of detection

7

High, Reduced primary function

performance, High degree of

customer dissatisfaction

Very low chance of detection

6

Moderate, Loss of secondary

function, partial malfunction of the

product - customer dissatisfied.

Low chance of detection

5

Low, Reduced secondary function

performance, will cause the

customers to complain

Moderate chance of detection

4Very Low, noticed by most

customers

Moderately high chance of

detection

3Minor effect, noticed by some

customers

High chance of detection

2

Very Minor, noticed by

discriminating customers

Very high chance of detection

1

No effect, unnoticed Remote likelihood of failure Current Control will almost

certainly detect / prevent Failure

Mode or Root Cause

Low: Relatively few failures

Moderate: Occasional Failure

High: Repeated Failures

Very high: Failure is almost

inevitable

Actual scale

should be

determined

based on

product type

and other risk

factors.


The output of an FMEA is the Risk Priority Number. The RPN is a calculated number based on the Severity, Occurrence, and Detection scoring:

of Effect of Root Cause / Failure

Mode / EffectEffectiveness of Control

RPN = Severity x Occurrence x Detection


Example: Pin (component of larger assy)

Jaws remain locked

on vesselInsufficient material


A B C D E F G H I J K

Ite

m #


Mode(s)

Potential Effect(s) of

Failure

Se

ve

rity

Potential Root

Cause(s)

Oc

cu

rre

nc

e

Current Design

Control(s)

Co

ntr

ol

sta

tus

De

tec

tio

n

RP

N

COMPONENT #1



3 0

4 0

COMPONENT #2



7 0

8 0

COMPONENT #3



11 0

12 0

COMPONENT #4



15 0

16 0


finish spec’d for PinActuates pawlExcessive friction,

pawl binds on pin

during use

Jaws remain locked

on vessel

Test to determine

functional life of

matl finish

PIN

9 3 542


Q: How do you determine what RPN level is actionable?

A: The preferred method is Pareto principle: Create a Pareto diagram of your

RPNs and address the top 10-30 percent first. Note that any Severity score of 9

or 10 should be considered for action even with a low Occurrence / Detection

score. The actionable RPN may be different for each product / project.

Sample

RPNs and

decisions.

Your

decisions

may be

different!

Sev Occ Det RPN Result Actions

1 2 1 2 Best Design Possible Actions not required.

2 2 10 40 A rare but undetectable failure with

little impact to the user

Action may be required. Determine if

detection / prevention can be improved

10 1 1 10 Catastrophic failure which rarely

happens and is detectable /

preventable

Action may be required. Review item to

determine if redesign is required to reduce

severity

10 2 10 200 Catastrophic and undetectable

failure which rarely reaches user

Action required. Improve detection or

redesign to reduce severity

3 10 2 60 Minor failure which occurs

frequently, effective prevention /

detection

Action required. Redesign to reduce

occurrence.

3 10 10 300 Minor frequent failure,

undetectable, unpreventable

Action required. Redesign / improve

detection

10 5 4 200 Catastrophic and relatively

frequent failure, marginally

effective controls

Action required. Redesign to reduce severity

10 10 10 1000 Catastrophic and frequent failure

with which will reach the user

Hold project

31ASQ NEQC 60th Conference, Springfield, MassachusettsOctober 22, 2014

R

P

N

Actions

RecommendedResp. Actions Taken

S

E

V

O

C

C

D

E

T

What are the actions

for reducing the

occurrence of the

Cause, or improving

Detection?

Who is

Responsible for

the recommended

action? And when

is it due?

What are the completed

actions taken with the

recalculated RPN? Be

sure to include

completion month/year

30No actions required

6 5 1

300

Material change to

high-strength

compound

J. Juran 3/1/14 Material changed 3/4/08,

testing completed - issue

corrected.

9 2 2

150Perform firing test

protocol RD-124

W. E. Deming

3/1/14

Test completed -

occurrence of issue unlikely 6 1 3

Completing Recommended Actions

Assign actions and

responsibilitiesRecord actual

actions taken

and when

completed

Re-score Severity*, Detection,

Occurrence and recalculate

RPN based on completed

actions

NOTE: Reducing the Severity is usually not

possible without a design change. An exception

may be if through testing, you realize that your

initial Severity score was excessive.


Tips for doing Better DFMEAs

33

Use as a Design Tool:

• “Bite-Sized” vs. Monolithic: A central tenant for

efficient product development is to improve flow by

reducing “batch” sizes.

• Tool vs. Checkbox: Design FMEA is not a

beneficial tool if used solely as a “check-box”

requirement.

• Tool vs. Tracker: A common misapplication of the tool is to use the

Design FMEA as a “tracker” to capture and track every risk mitigation

action.

• Early and Throughout vs. Last Minute: A Good DFMEA encourages

brainstorming potential issues early in the design cycle and improving

the design throughout testing and evaluation.



34 |

Efficient and Intuitive:

• Who?: Team effort vs. Designer’s Responsibility: A Good DFMEA

might consider a more balanced approach to individual responsibilities

and team meeting.

• Appropriate vs. Overly-Detailed: A Good FMEA does not

necessarily break everything down to the component level

• Non-Intuitive vs. “Oh, this make sense!”:

• “Chain of Events” Thinking

• Scoring Afterwards

• Replacing “Occurrence” with “Likelihood”

• Remove Detection

• Reduced Scales

October 22, 2014ASQ NEQC 60th Conference,

Springfield, Massachusetts


35

Leverage a Facilitator:

• Pre-work. Stay ahead of the team. Complete

the framework for the meeting agenda ahead

of time.

• Agenda: Have an agenda and state it at the

beginning of each meeting:

• Clear Objectives: Write the high level meeting objectives in bold letters

on white board or flip chart.

• No open laptop computers in the meeting, other than the facilitator’s:

Phones and BlackBerrys put away. Do not waste the group’s time typing

and formatting to perfection.

• Visual Room: Place definitions of key terms as well as criteria charts for

Severity and Likelihood on the wall.

• JIT Training: Make sure the group understands the basic terminology

and “chain of events” concept before brainstorming.



36

Leverage a Facilitator:

• Brainstorm first: Then cover Current Controls, and then score/rank.

• Focus: Keep the conversation on track.

• Engaged Participants: Call on participants by name with a pertinent

question if they have not contributed in a while.

• Have Fun: Candy helps – it is important to keep the team positive.

• Administration: Keep a parking lot to capture off topic action items and

take five minutes at the end of the meeting getting feedback on the

overall process.

• Parts: Bring examples of the

components / product or similar products

to encourage brainstorming.


FMEA and Fault Tree Analysis

FMEA and Fault Tree Analysis complement each other:

FMEA starts at the component level failure and works upward to the product failure.

Fault Tree Analysis starts at the product failures and works downward into the causes. FTA is more often used for failure analysis of complex equipment.

37

FMEA: Upward

direction of analysis.

FTA: Downward

direction of analysis.

cause cause

Product

Failure

cause

causecause

cause cause

Product

Failure

Systems

Components


38

Which is the better tool?

Condition FTA FMEA

Primary concern is safety X

A small number of top events X

Top events undefined or large number of top events X

Functional profile is of critical importance X

Identification of "all possible" failure modes X

High potential for human error X

High potential for software error X

Quantified risk evaluation X

Product functin is highly complex X

In many cases, the use of both techniques is appropriate.


1. Develop a block diagram of the Product / Component / Function being analyzed. For complex products, a breakdown to sub-system may be adequate.

2. Enter the Components and functions into the DFMEA template. Each component (or subsystem) function will be analyzed as a separate line item.

3. For each Component Function, determine the ways in which it can fail (Failure Modes). In some cases, it may be easier to brainstorm Root Causes first.

4. For each Failure Mode associated with the item, determine the Effects of the failure on the customer. Remember that a Customer may be internal, external, or an end user,.

5. Identify potential Root Causes of each Failure Mode.

6. List the Current Controls for each Root Cause or Failure Mode.

7. Create Severity, Occurrence, and Detection rating scales.

8. Assign Severity, Occurrence and Detection ratings to each line item.

9. Calculate RPN’s for each Root Cause.

10. Determine Recommended Actions to reduce high RPNs.

11. Take appropriate actions and recalculate RPNs.

12. Review AFMEA and PFMEA to determine if additional updates to those documents are required.


DFMEA Summary

40

Recommended

Action(s)

Responsibility and

Target Date

Actions Taken and

Implementation Date

Re

v S

ev

eri

ty

Re

v O

cc

ur.

Re

v D

ete

cti

on

Re

v R

PN

Ite

m #


Mode(s)

Potential Effects of

Failure

Se

ve

rity

Potential Causes /

Mechanism of

Failure

Oc

cu

rre

nc

e

Current Design

Control(s)

De

tec

tio

n

RP

N

What is the

function?

HOW can thefailure occur?

What are the effects

of the failure?

How bad is thefailure - assuming

it occurs?

What is the root cause

of the failure?

How likely is theRoot Cause / Failure

Mode / Effect?

What do we

currently do to

prevent the failure

or enable detection

of imminent failure?

How effective is

the Current

Control?

What's theoverall risk?

What do we

need to do?

Who's

doing it and

when?

What did we actually

do?

Did we make

the effect

less severe?

Did we make the

Potential Failure less

likely?

Did we prevent the

failure or improve

detectability?

Did we reducethe overall risk?



Thank you

Questions ?


[email protected]

mailto:[email protected]

Good FMEA / Bad FMEA William... · · 2016-06-18Good FMEA / Bad FMEA By: William Mulligan Session 23E. ... the Apollo missions. 1970s • MIL-STD-1629 ... Material softens with

Documents