fmea.pdf

FMEA

FMEA

S4 /IEE Application Examples: FMEA

Implementation

Development of a Design FMEA

Design FMEA Tabular Entries

Development of a Process FMEA

Process FMEA Tabular Entries

FMEA

To remain competitive, organizations must continually improve. Potential fail-ure mode and effects analysis (FMEA) is a method that facilitates processimprovement. Using FMEAs, organizations can identify and eliminate con-cerns early in the development of a process or design and provide a form ofrisk analysis. The quality of procured parts or services can improve whenorganizations work with their suppliers to implement FMEAs within theirorganization. Properly executed FMEAs can improve internal and externalcustomer satisfaction in addition to the bottom line of organizations.

Discussed in this chapter are design and process FMEAs. Design FMEA(DFMEA) applications include component, subsystem, and main system. Pro-cess FMEA (PFMEA) applications include assembly, machines, workstations,gages, procurement, training of operators, and tests.

Benefits of a properly executed FMEA include:

• Improved product functionality and robustness• Reduced warranty costs• Reduced day-to-day manufacturing problems• Improved safety of products and implementation processes• Reduced business process problems

S 4/IEE APPLICATION EXAMPLES: FMEA

S4/IEE application examples of FMEA are:

• Transactional 30,000-foot-level metric: DSO reduction was chosen as anS4/IEE project. The team used a cause-and-effect matrix to prioritizeitems from a cause-and-effect diagram. An FMEA was conducted of theprocess steps and/or highest categories from the cause-and-effect matrix.

• Manufacturing 30,000-foot-level metric (KPOV): An S4/IEE project wasto improve the capability/performance of a process that affected the di-ameter of a manufactured product (i.e., reduce the number of parts be-yond the specification limits). The team used a cause-and-effect matrixto prioritize items from a cause-and-effect diagram. An FMEA was con-ducted of the process steps and/or highest categories from the cause-and-effect matrix.

• Transactional and manufacturing 30,000-foot-level cycle time metric (alean metric): An S4/IEE project was to improve the time from order entryto fulfillment was measured. The team used a cause-and-effect matrix toprioritize items from a cause-and-effect diagram. An FMEA was con-ducted of the process steps and/or highest categories from the cause-and-effect matrix.

• Transactional and manufacturing 30,000-foot-level inventory metric orsatellite-level TOC metric (a lean metric): An S4/IEE project was toreduce inventory. The team used a cause-and-effect matrix to prioritizeitems from a cause-and-effect diagram. An FMEA was conducted of theprocess steps and/or highest categories from the cause-and-effect matrix.

• Manufacturing 30,000-foot-level quality metric: An S4/IEE project wasto reduce the number of defects in a printed circuit board manufacturingprocess. The team used a cause-and-effect matrix to prioritize items froma cause-and-effect diagram. An FMEA was conducted of the processsteps and/or highest categories from the cause-and-effect matrix.

• Product DFSS: An S4/IEE product DFSS project was to reduce the30,000-foot-level metric of number of product phone calls generated fornewly developed products. The team used a cause-and-effect matrix toprioritize items from a cause-and-effect diagram. An FMEA was con-ducted of the process steps when developing a product and/or highestcategories from the cause-and-effect matrix. One process-improvementidea for the development process was to establish a product designFMEA procedure.

• Process DFSS: A team was to create a new call center. A process flow-chart of the planned call center process was created. An FMEA wasconducted to assess risks for steps within this process and then createaction plans to address identified issues.

IMPLEMENTATION

Timeliness and usefulness as a living document are important aspects of asuccessful FMEA. To achieve maximum benefit, organizations need to con-duct FMEAs before a failure is unknowingly instituted into a process ordesign.

FMEA input is a team effort, but one individual typically is responsible,by necessity, for its preparation. It is the role of the responsible engineer toorchestrate the active involvement of representatives from all affected areas.FMEAs should be part of design or process concept finalization that acts asa catalyst for the stimulation and interchange of ideas between functions. AnFMEA should be a living document that is updated for design changes andthe addition of new information.

Important FMEA implementation issues include the following:

• Use as a living document with periodic review and updates.• Conduct early enough in development cycle to

• Design out potential failure modes by eliminating root causes.• Reduce seriousness of failure mode if elimination is not possible.• Reduce occurrence of the failure mode.

Implementation benefits of an FMEA include the following:

• Early actions in the design cycle save time and money.• Thorough analysis with teams creates better designs and processes.• Complete analysis provides possible legal evidence.• Previous FMEAs provide knowledge leading to current design or product

FMEAs.

Team interaction is an important part of executing an FMEA. Organizationsshould consider using outside suppliers in an FMEA and creating the teamso that it consists of five to seven knowledgeable, active members. Whenexecuting an FMEA, teams work to identify potential failure modes for designfunctions or process requirements. They then assign a severity to the effectof this failure mode. They also assign a frequency of occurrence to the po-tential cause of failure and likelihood of detection. Organizations can differin approach to assigning numbers to these factors (i.e., severity, frequency ofoccurrence, and likelihood of detection—sometimes called SOD values), withthe restriction that higher numbers are worse. After these numbers are deter-mined, teams calculate a risk priority number (RPN), which is the product ofthese three numbers. Teams use the ranking of RPNs to focus process im-provement efforts.

An effective roadmap to create FMEA entries is as follows:

• Note an input to a process or design (e.g., process step, key input iden-tified in a cause-and-effect matrix, or design function).

• List two or three ways input/function can go wrong.• List at least one effect of failure.• For each failure mode, list one or more causes of input going wrong.• For each cause list at least one method of preventing or detecting cause.• Enter SOD values.

DEVELOPMENTOFADESIGNFMEA

Within a design FMEA, manufacturing and/or process engineering input isimportant to ensure that the process will produce to design specifications. Ateam should include knowledgeable representation from design, test, reliabil-ity, materials, service, and manufacturing/process organizations.

A design FMEA presumes the implementation of manufacturing/assemblyneeds and design intents. A design FMEA does not need to include potentialfailure modes, causes, and mechanisms originating from manufacturing/as-sembly when their identification, effect, and control is covered by a processFMEA. However, a design FMEA team may choose to consider some processFMEA issues. Design FMEAs do not rely on process controls to overcomepotential design weaknesses, but do consider technical and physical limits ofthe manufacturing/assembly process.

When beginning a design FMEA, the responsible design engineer compilesdocuments that give insight into the design intent. Design intent is expressedas a list of what the design is expected to do and what it is not expected todo. Quality function deployment (QFD) and manufacturing/assembly require-ments are sources for determining the design wants and needs of customers.The identification of potential failure modes for corrective action is easiestwhen the design intent is clear and thorough.

A block diagram of the system, subsystem, and/or component at the be-ginning of a design FMEA is useful to improve the understanding of the flowof information and other characteristics for the FMEA. Blocks are the func-tions, while the deliverables are the inputs and outputs of the blocks. Theblock diagram shows the relationship between analysis items and establishesa logical order for analysis. The documentation for an FMEA should includeits block diagram. Figure 14.1 provides one example of a relational blockdiagram; other types of block diagrams may be more useful, depending onthe specific items considered in the analysis.

Table 14.1 shows a blank FMEA form. A team determines the designFMEA tabular entries following guidelines as described in the next section.

FIGURE 14.1 Relational block diagram example. [Reprinted with permission fromthe FMEA Manual (Chrysler, Ford, General Motors Supplier Quality RequirementsTask Force).]

TA

BL

E14

.1B

lank

FM

EA

For

m

POT

EN

TIA

L

FAIL

UR

EM

OD

EA

ND

EFF

EC

TS

AN

ALY

SIS

FME

ATy

pe(D

esig

nor

Proc

ess)

:Pr

ojec

tN

ame

/Des

crip

tion:

Dat

e(O

rig.

):

Res

pons

ibili

ty:

Prep

ared

By:

Dat

e(R

ev.)

:

Cor

eTe

am:

Dat

e(K

ey):

Des

ign

FME

A(I

tem

/Fu

nctio

n)Pr

oces

sFM

EA

(Fun

ctio

n/

Req

uire

men

ts)

Pote

ntia

lFa

ilure

Mod

e

Pote

ntia

lE

ffec

t(s)

ofFa

ilure

S e v

C l a s s

Pote

ntia

lC

ause

(s)/

Mec

hani

sm(s

)of

Failu

re

O c c u r

Cur

rent

Con

trol

sPr

even

tion

Cur

rent

Con

trol

sD

etec

tion

D e t e c

R P NR

ecom

men

ded

Act

ions

Res

pons

ibili

ty&

Targ

etC

ompl

etio

nD

ate

Act

ions

Take

n

S e v

O c c u r

D e t e c

R P N

DESIGNFMEATABULARENTRIES

A design FMEA in the format of Table 14.1 contains the following:

• Header Information. Documents the system/subsystem/component (un-der project name/description) and supplies other information about whenand who created the FMEA.

• Item/Function. Contains the name and number of the item to be ana-lyzed. Includes a concise, exact, and easy-to-understand explanation ofa function of the item task or response that is analyzed to see whetherit meets the intent of the design. Includes information regarding the tem-perature, pressure, and other pertinent system operating conditions. Whenthere is more than one function, it lists each function separately, withdifferent potential failure modes.

• Potential Failure Mode. Describes ways a design might fail to performits intended function. May include the cause of a potential failure modein a higher-level subsystem or process step. May also be the effect of afailure in a lower-level component or process step. Contains, for eachitem/function, a list of potential failure modes given the assumption thatthe failure might occur. Items considered are previous problems and newissues from brainstorming sessions. Consideration is given to problemsthat could arise only under certain operation conditions, such as hightemperature and high humidity. Descriptions are in physical or technicalterms, not symptoms. Includes failure modes such as fractured, electricalshort-circuited, oxidized, and circuit logic failed.

• Potential Effect(s) of Failure. Describes the effects of the failure modeon the function from an internal or external customer point of view.Highlights safety or noncompliance with regulation issues. Expressed interms of the specific system, subsystem, or component hierarchical re-lationship that is analyzed. Includes failure effects such as intermittentoperation, lost computer data, and poor performance.

• Severity. Assesses the seriousness of the effect of the potential failuremode to the next component, subsystem, or system. Design change usu-ally strives to reduce severity levels. Estimation is typically based on a1 to 10 scale where the team agrees to a specific evaluation criteria foreach ranking value. Table 14.2 shows example evaluation criteria for theautomotive industry.

• Classification. Includes optional information such as critical character-istics requiring additional process controls. An appropriate character orsymbol in this column indicates the need for an entry in the recom-mended action column and special process controls within the processFMEA.

• Potential Causes(s) of Failure. Indicates a design weakness that causesthe potential failure mode. Contains a concise, clear, and comprehensive

TABLE 14.2 Severity Evaluation Criterion Example for Design FMEA

Effect Criteria: Severity of Effect Ranking

Hazardouswithoutwarning

Very high severity ranking when a potential failuremode affects safe vehicle operation and/orinvolves noncompliance with governmentregulations without warning.

10

Hazardouswithwarning

Very high severity ranking when a potential failuremode affects safe vehicle operation and/orinvolves noncompliance with governmentregulation with warning.

9

Very high Vehicle / item inoperable (loss of primary function). 8High Vehicle / item operable, but at reduced level of

performance. Customer very dissatisfied.7

Moderate Vehicle / item operable, but comfort /convenienceitem(s) inoperable. Customer dissatisfied.

6

Low Vehicle / item operable, but comfort /convenienceitem(s) operable at reduced level of performance.Customer somewhat dissatisfied.

5

Very low Fit & finish/squeak & rattle item does notconform. Defect noticed by most customers(greater than 75%).

4

Minor Fit & finish/squeak & rattle item does notconform. Defect noticed by 50% of customers.

3

Very minor Fit & finish/squeak & rattle item does notconform. Defect noticed by discriminatingcustomers (less than 25%).

2

None No discernible effect. 1

Source: Reprinted with permission from the FMEA Manual (DaimlerChrysler, Ford Motor Com-pany, General Motors Supplier Quality Requirements Task Force).

list of all root causes (not symptoms) of failure. Includes causes such asincorrect algorithm, hardness, porosity, and incorrect material specified.Includes failure mechanisms such as fatigue, wear, and corrosion.

• Occurrence. Estimates the likelihood that a specific cause will occur.Consideration of historical data of components/subsystems similar to thenew design helps determine the ranking value. Teams need to agree onan evaluation criterion, where possible failure rates are anticipated valuesduring design life. Table 14.3 shows example occurrence criteria.

• Current Design Controls. Lists activities such as design verification tests,design reviews, DOEs, and tolerance analysis that ensure adequacy ofdesign control for the failure mode. In an update to their booklet, AIAG(2001) changed this from a one-column category to a two-column cate-gory, where one column is for prevention, while the other column is fordetection.

TABLE 14.3 Occurrence Evaluation Criterion Example for Design FMEA

Probability of Failure Possible Failure Rates Ranking

Very high: Persistent failures �100 per thousand vehicles / items50 per thousand vehicles / items

109

High: Frequent failures 20 per thousand vehicles / items10 per thousand vehicles / items

87

Moderate: Occasional failures 5 per thousand vehicles / items2 per thousand vehicles / items1 per thousand vehicles / items

654

Low: Relatively few failures 0.5 per thousand vehicles / items0.1 per thousand vehicles / items

32

Remote: Failure is unlikely �0.010 per thousand vehicles / items 1


• Detection. Assessment of the ability of the current design control to de-tect the subsequent failure mode or potential cause of design weaknessbefore releasing to production. Table 14.4 shows example detection cri-teria.

• Risk Priority Number (RPN). Product of severity, occurrence, and detec-tion rankings. The ranking of RPN prioritizes design concerns; however,problems with a low RPN still deserve special attention if the severityranking is high.

• Recommended Action(s). This entry proposes actions intended to lowerthe occurrence, severity, and/or detection rankings of the highest RPNfailure modes. Example actions include DOE, design revision, and testplan revision. ‘‘None’’ indicates that there are no recommended actions.

• Responsibility for Recommended Action. Documents the organization andindividual responsible for recommended action and target completiondate.

• Actions Taken. Describes implementation of recommended action andeffective date.

• Resulting RPN. Contains the recalculated RPN resulting from correctiveactions that affected previous severity, occurrence, and detection rank-ings. Blanks indicate no action taken.

The responsible design engineer follows up to ensure the adequate imple-mentation of all recommended actions. An FMEA should include designchanges and other relevant actions, even after the start of production. Table14.5 exemplifies a completed process FMEA.

TABLE 14.4 Detection Evaluation Criterion Example for Design FMEA

DetectionCriteria: Likelihood of Detection

by Design Control Ranking

Absolute uncertainty Design control will not and/or cannot detect apotential cause/mechanism and subsequentfailure mode; or there is no design control.

10

Very remote Very remote chance the design control willdetect a potential cause/mechanism andsubsequent failure mode.

9

Remote Remote chance the design control will detect apotential cause/mechanism and subsequentfailure mode.

8

Very low Very low chance the design control will detect apotential cause/mechanism and subsequentfailure mode.

7

Low Low chance the design control will detect apotential cause/mechanism and subsequentfailure mode.

6

Moderate Moderate chance the design control will detecta potential cause/mechanism and subsequentfailure mode.

5

Moderately high Moderately high chance the design control willdetect a potential cause/mechanism andsubsequent failure mode.

4

High High chance the design control will detect apotential cause/mechanism and subsequentfailure mode.

3

Very high Very high chance the design control will detecta potential cause/mechanism and subsequentfailure mode.

2

Almost certain Design control will almost certainly detect apotential cause/mechanism and subsequentfailure mode.

1


DEVELOPMENTOFAPROCESSFMEA

For a process or assembly FMEA, design engineering input is important toensure the appropriate focus on important design needs. An effort of the teamshould include knowledgeable representation from design, manufacturing/process, quality, reliability, tooling, and operators.

A process FMEA presumes the product meets the intent of the design. Aprocess FMEA does not need to include potential failure modes, causes, and

TA

BL

E14

.5E

xam

ple:

Pot

enti

alF

ailu

reM

ode

and

Eff

ects

Ana

lysi

s(D

esig

nF

ME

A)

Syst

emSu

bsys

tem

x

Com

pone

nt01

.03

/Bod

yC

losu

res

Des

ign

Res

pons

ibili

tyB

ody

Eng

inee

ring

FME

AN

umbe

r12

34

Page

1of

1

Prep

ared

By

A.

Tate

—X

6412

—B

ody

Eng

inee

r

Mod

elY

ear(

s)/V

ehic

le(s

)19

9X/L

ion

4do

or/W

agon

Key

Dat

e9X

0301

ER

FME

AD

ate

(Ori

g.)

8X03

22(R

ev.)

8X07

14

Item

Func

tion

Pote

ntia

lFa

ilure

Mod

ePo

tent

ial

Eff

ect(

s)of

Failu

re

S e v

C l a s sPo

tent

ial

Cau

se(s

)/M

echa

nism

(s)

ofFa

ilure

O c c u r

Cur

rent

Des

ign

Con

trol

sPr

even

tion

Cur

rent

Des

ign

Con

trol

sD

etec

tion

D e t e c

R P NR

ecom

men

ded

Act

ion(

s)

Res

pons

ibili

tyan

dTa

rget

Com

plet

ion

Dat

eA

ctio

nsTa

ken

S e v

O c c

D e t

R P N

Fron

tdo

orL

.H.

H8H

X-0

000-

A

•In

gres

sto

and

egre

ssfr

omve

hicl

e•

Occ

upan

tpr

otec

tion

from

wea

ther

,no

ise,

and

side

impa

ct•

Supp

ort

anch

orag

efo

rdo

orha

rdw

are

incl

udin

gm

irro

r,hi

nges

,la

tch

and

win

dow

regu

lato

r•

Prov

ide

prop

ersu

rfac

efo

rap

pear

ance

item

s•

Pain

tan

dso

fttr

im

Cor

rode

din

teri

orlo

wer

door

pane

ls

Det

erio

rate

dlif

eof

door

lead

ing

to:

•U

nsat

isfa

ctor

yap

pear

ance

due

toru

stth

roug

hpa

int

over

time

•Im

pair

edfu

nctio

nof

inte

rior

door

hard

war

e

7U

pper

edge

ofpr

otec

tive

wax

appl

icat

ion

spec

ified

for

inne

rdo

orpa

nels

isto

olo

w

Insu

ffici

ent

wax

thic

knes

ssp

ecifi

ed

Inap

prop

riat

ew

axfo

rmul

atio

nsp

ecifi

ed

Ent

rapp

edai

rpr

even

tsw

axfr

omen

teri

ngco

rner

/edg

eac

cess

6 4 2 5

Veh

icle

gene

ral

dura

bilit

yte

stva

h.T-

118

T-10

9T-

301

Veh

icle

gene

ral

dura

bilit

yte

stin

g(a

sab

ove)

Phys

ical

and

Che

mL

abte

st:

Rep

ort

No.

1265

Des

ign

aid

inve

stig

atio

nw

ithno

nfun

ctio

ning

spra

yhe

ad

7 7 2 8

294

196 28 280

Add

labo

rato

ryac

cele

rate

dco

rros

ion

test

ing

Add

labo

rato

ryac

cele

rate

dco

rros

ion

test

ing

Con

duct

desi

gnof

expe

rim

ents

(DO

E)

onw

axth

ickn

ess

Non

e

Add

team

eval

uatio

nus

ing

prod

uctio

nsp

ray

equi

pmen

tan

dsp

ecifi

edw

ax

ATa

te-B

ody

Eng

inee

ring

8X09

30

Com

bine

w/t

est

for

wax

uppe

red

geve

rific

atio

n

ATa

tebo

dyen

gine

erin

g9X

0115

Bod

yen

gine

erin

gan

das

sem

bly

oper

atio

ns8X

1115

Bas

edon

test

resu

lts(t

est

no.

1481

)up

per

edge

spec

rais

ed12

5m

m

Test

resu

lts(t

est

no.

1481

)sh

owsp

ecifi

edth

ickn

ess

isad

equa

te.

DO

Esh

ows

25%

vari

atio

nin

spec

ified

thic

knes

sis

acce

ptab

le.

7 7 7

2 2 1

2 2 3

28 28 21

Wax

appl

icat

ion

plug

sdo

ordr

ain

hole

s3

Lab

orat

ory

test

usin

g‘‘

wor

st-c

ase’

’wax

appl

icat

ion

and

hole

size

121

Non

eB

ased

onte

st,

thre

ead

ditio

nal

vent

hole

spr

ovid

edin

affe

cted

area

s

Insu

ffici

ent

room

betw

een

pane

lsfo

rsp

ray

head

acce

ss

4D

raw

ing

eval

uatio

nof

spra

yhe

adac

cess

411

2A

ddte

amev

alua

tion

usin

gde

sign

aid

buck

and

spra

yhe

ad

Bod

yen

gine

erin

gan

das

sem

bly

oper

atio

ns

Eva

luat

ion

show

edad

equa

teac

cess

71

17

SAM

PLE

Sour

ce:

Rep

rint

edw

ithpe

rmis

sion

from

the

FM

EA

Man

ual

(Dai

mle

rChr

ysle

r,Fo

rdM

otor

Com

pany

,G

ener

alM

otor

sSu

pplie

rQ

ualit

yR

equi

rem

ents

Task

Forc

e).

mechanisms originating from the design, though a process FMEA team maychoose to include some design issues. The design FMEA covers the effectand avoidance of these issues. A process FMEA can originate from a flow-chart that identifies the characteristics of the product/process associated witheach operation. Included are appropriate product effects from available designFMEA. The documentation for an FMEA should include its flowchart.

Table 14.1 shows a blank FMEA form. A team determines the processFMEA tabular entries following the guidelines presented in the next section.

PROCESSFMEATABULARENTRIES

A process FMEA in the format of Table 14.1 contains the following:

• Header Information. Documents the process description and suppliesother information about when and who created the FMEA.

• Process Function/Requirements from a Process FMEA. Contains a sim-ple description of the process or operation analyzed. Example processesinclude assembly, soldering, and drilling. Concisely indicates the purposeof the analyzed process or operation. When numeric assembly operationsexist with differing potential failure modes, the operations may be listedas separate processes.

• Potential Failure Mode. Describes how the process could potentially failto conform to process requirements and/or design intent at a specificoperation. Contains for each operation or item/function a list of eachpotential failure mode in terms of the component, subsystem, system, orprocess characteristic. Consider how the process/part fails to meet spec-ifications and/or customer expectations. Subsequent or previous opera-tions can cause these failure modes; however, teams should assume thecorrectness of incoming parts and materials. Items considered are pre-vious problems and new issues foreseen by brainstorming. Includes fail-ure modes such as broken, incorrect part placement, and electricalshort-circuited.

• Potential Effect(s) of Failure. Describes the effects of the failure modeon the function from an internal or external customer point of view.Considers what the customer experiences or the ramifications of thisfailure mode either from the end-user point of view or from subsequentoperation steps. Example end-user effects are poor performance, inter-mittent failure, and poor appearance. Example subsequent operation ef-fects are ‘‘does not fit,’’ ‘‘cannot mount,’’ and ‘‘fails to open.’’

• Severity. Assesses the seriousness of the effect of the potential failuremode to the customer. Estimation is typically based on a 1 to 10 scalewhere the team agrees to a specific evaluation criterion for each rankingvalue. Table 14.6 shows example evaluation criterion for the automotiveindustry.

TA

BL

E14

.6Se

veri

tyE

valu

atio

nC

rite

rion

Exa

mpl

efo

rP

roce

ssF

ME

A

Eff

ect

Cri

teri

a:Se

veri

tyof

Eff

ect

Thi

sra

nkin

gre

sults

whe

na

pote

ntia

lfa

ilure

mod

ere

sults

ina

final

cust

omer

and

/or

am

anuf

actu

ring

/ass

embl

ypl

ant

defe

ct.

The

final

cust

omer

shou

ldal

way

sbe

cons

ider

edfir

st.

Ifbo

thoc

cur,

use

the

high

erof

the

two

seve

ritie

s.

Cus

tom

erE

ffec

tM

anuf

actu

ring

/Ass

embl

yE

ffec

tR

anki

ng

Haz

ardo

usw

ithou

tw

arni

ng

Ver

yhi

ghse

veri

tyra

nkin

gw

hen

apo

tent

ial

failu

rem

ode

affe

cts

safe

vehi

cle

oper

atio

nan

d/o

rin

volv

esno

ncom

plia

nce

with

gove

rnm

ent

regu

latio

nw

ithou

tw

arni

ng.

Or

may

enda

nger

oper

ator

(mac

hine

oras

sem

bly)

with

out

war

ning

.10

Haz

ardo

usw

ithw

arni

ng

Ver

yhi

ghse

veri

tyra

nkin

gw

hen

apo

tent

ial

failu

rem

ode

affe

cts

safe

vehi

cle

oper

atio

nan

d/o

rin

volv

esno

ncom

plia

nce

with

gove

rnm

ent

regu

latio

nw

ithw

arni

ng.

Or

may

enda

nger

oper

ator

(mac

hine

oras

sem

bly)

with

war

ning

.9

Ver

yhi

ghV

ehic

le/i

tem

inop

erab

le(l

oss

ofpr

imar

yfu

nctio

n).

Or

100%

ofpr

oduc

tm

ayha

veto

besc

rapp

ed,

orve

hicl

e/ite

mre

pair

edin

repa

irde

part

men

tw

itha

repa

irtim

egr

eate

rth

anon

eho

ur.

8

Hig

hV

ehic

le/i

tem

oper

able

but

ata

redu

ced

leve

lof

perf

orm

ance

.C

usto

mer

very

diss

atis

fied.

Or

prod

uct

may

have

tobe

sort

edan

da

port

ion

(les

sth

an10

0%)

scra

pped

,or

vehi

cle/

item

repa

ired

inre

pair

depa

rtm

ent

with

are

pair

time

betw

een

aha

lf-h

our

and

anho

ur.

7

Mod

erat

eV

ehic

le/i

tem

oper

able

but

com

fort

/con

veni

ence

item

(s)

inop

erab

le.

Cus

tom

erdi

ssat

isfie

d.O

ra

port

ion

(les

sth

an10

0%)

ofth

epr

oduc

tm

ayha

veto

besc

rapp

edw

ithno

sort

ing,

orve

hicl

e/ite

mre

pair

edin

repa

irde

part

men

tw

itha

repa

irtim

ele

ssth

ana

half

-hou

r.

6

Low

Veh

icle

/ite

mop

erab

lebu

tco

mfo

rt/c

onve

nien

ceite

m(s

)op

erab

leat

redu

ced

leve

lof

perf

orm

ance

.O

r10

0%of

prod

uct

may

have

tobe

rew

orke

d,or

vehi

cle/

item

repa

ired

off-

line

but

does

not

goto

repa

irde

part

men

t.

5

Ver

ylo

wFi

tan

dfin

ish

/squ

eak

and

rattl

eite

mdo

esno

tco

nfor

m.

Def

ect

notic

edby

mos

tcu

stom

ers

(gre

ater

than

75%

).

Or

the

prod

uct

may

have

tobe

sort

ed,

with

nosc

rap,

and

apo

rtio

n(l

ess

than

100%

)re

wor

ked.

4

Min

orFi

tan

dfin

ish

/squ

eak

and

rattl

eite

mdo

esno

tco

nfor

m.

Def

ect

notic

edby

50%

ofcu

stom

ers.

Or

apo

rtio

n(l

ess

than

100%

)of

the

prod

uct

may

have

tobe

rew

orke

d,w

ithno

scra

p,on

-lin

ebu

tou

t-of

-sta

tion.

3

Ver

ym

inor

Fit

and

finis

h/s

quea

kan

dra

ttle

item

does

not

conf

orm

.D

efec

tno

ticed

bydi

scri

min

atin

gcu

stom

ers

(les

sth

an25

%).

Or

apo

rtio

n(l

ess

than

100%

)of

the

prod

uct

may

have

tobe

rew

orke

d,w

ithno

scra

p,on

-lin

ebu

tin

-st

atio

n.

2

Non

eN

odi

scer

nibl

eef

fect

.O

rsl

ight

inco

nven

ienc

eto

oper

atio

nor

oper

ator

,or

noef

fect

.1

Sour

ce:

Rep

rint

edw

ithpe

rmis

sion

from

the

FM

EA

Man

ual

(Dai

mle

rChr

ysle

r,Fo

rdM

otor

Com

pany

,G

ener

alM

otor

sSu

pplie

rQ

ualit

yR

equi

rem

ents

Task

Forc

e).

TABLE 14.7 Occurrence Evaluation Criterion Example for Process FMEA

Probability Likely Failure Rates Ranking

Very high: Persistent failures �100 per thousand pieces50 per thousand pieces

109

High: Frequent failures 20 per thousand pieces10 per thousand pieces

87

Moderate: Occasional failures 5 per thousand pieces2 per thousand pieces1 per thousand pieces

654

Low: Relatively few failures 0.5 per thousand pieces0.1 per thousand pieces

32

Remote: Failure unlikely �0.01 per thousand pieces 1


• Classification. Includes optional information that classifies special pro-cess characteristics that may require additional process controls. Applieswhen government regulations, safety, and engineering specificationconcerns exist for the product and/or process. An appropriate charac-ter or symbol in this column indicates the need for an entry in therecommended action column to address special controls in the controlplan.

• Potential Causes(s) of Failure. Describes how failure could occur interms of a correctable or controllable item. Contains a concise, descrip-tive, and comprehensive list of all root causes (not symptoms) of failure.The resolution of some causes directly affects the failure mode. In othersituations a DOE determines the major and most easily controlled rootcauses. Includes causes such human error, improper cure time, and miss-ing part.

• Occurrence. Estimates the frequency of occurrence of failure withoutconsideration of detecting measures. Gives the number of anticipatedfailures during the process execution. Consideration of statistical datafrom similar processes improves the accuracy of ranking values. Alter-native subjective assessments use descriptive words to describe rankings.Table 14.7 shows example occurrence criteria.

• Current Process Controls. Describes controls that can prevent failuremode from occurring or detect occurrence of the failure mode. In anupdate to their booklet, AIAG (2001) changed this from a one-columncategory to a two-column category, where one column is for prevention,while the other column is for detection. Process controls includes controlmethods such as SPC and poka-yoke (fixture error proofing) at the sub-

ject or subsequent operations. The preferred method of control is pre-vention or reduction in the frequency of the cause/mechanism to thefailure mode/effect. The next preferred method of control is detection ofthe cause/mechanism, which leads to corrective actions. The least pre-ferred method of control is detection of the failure mode.

• Detection. Assesses the probability of detecting a potential cause/mech-anism from process weakness or the subsequent failure mode before thepart /component leaves the manufacturing operation. Ranking values con-sider the probability of detection when failure occurs. Table 14.8 showsexample detection evaluation criteria.

• Risk Priority Number (RPN). Product of severity, occurrence, and detec-tion rankings. The ranking of RPN prioritizes design concerns; however,problems with a low RPN still deserve special attention if the severityranking is high.

• Recommended Action(s). This entry is proposed actions intended to lowerthe occurrence, severity, and/or detection rankings of the highest RPNfailure modes. Example actions include DOE to improve the understand-ing of causes and control charts to improve the focus of defectprevention/continuous improvement activities. Teams should focus onactivities that lead to the prevention of defects (i.e., occurrence rankingreduction) rather than improvement of detection methodologies (i.e., de-tection ranking reduction). Teams should implement corrective action toidentified potential failure modes where the effect is a hazard tomanufacturing/assembly personnel. Severity reduction requires a revisionin the design and/or process. ‘‘None’’ indicates that there are no rec-ommended actions.

• Responsibility for Recommended Action. Documents the organization andindividual responsible for recommended action and target completiondate.

• Actions Taken. Describes implementation of recommended action andeffective date.

• Resulting RPN. Contains the recalculated RPN resulting from correctiveactions that affected previous severity, occurrence, and detection rank-ings. Blanks indicate no action taken.

The responsible process engineer follows up to ensure the adequate imple-mentation of all recommended actions. An FMEA should include designchanges and other relevant actions even after the start of production. Table14.9 provides an example of a completed process FMEA that has an RPNtrigger number of 150, along with a severity trigger number of 7. Table 14.10is another example of a completed FMEA with an action RPN trigger numberof 130.

TA

BL

E14

.8D

etec

tion

Eva

luat

ion

Cri

teri

aE

xam

ple

for

Pro

cess

FM

EA

Det

ectio

nC

rite

ria

Insp

ectio

nTy

pe

AB

CSu

gges

tion

Ran

geof

Det

ectio

nM

etho

dsR

anki

ng

Alm

ost

impo

ssib

leA

bsol

ute

cert

aint

yof

nond

etec

tion.

XC

anno

tde

tect

oris

not

chec

ked.

10

Ver

yre

mot

eC

ontr

ols

will

prob

ably

not

dete

ct.

XC

ontr

olis

achi

eved

with

indi

rect

orra

ndom

chec

kson

ly.

9

Rem

ote

Con

trol

sha

vepo

orch

ance

ofde

tect

ion.

XC

ontr

olis

achi

eved

with

visu

alin

spec

tion

only

.8

Ver

ylo

wC

ontr

ols

have

poor

chan

ceof

dete

ctio

n.X

Con

trol

isac

hiev

edw

ithdo

uble

visu

alin

spec

tion

only

.7

Low

Con

trol

sm

ayde

tect

.X

XC

ontr

olis

achi

eved

with

char

ting

met

hods

,su

chas

SPC

(Sta

tistic

alPr

oces

sC

ontr

ol).

6

Mod

erat

eC

ontr

ols

may

dete

ct.

XC

ontr

olis

base

don

vari

able

gaug

ing

afte

rpa

rts

have

left

the

stat

ion,

orG

o/N

oG

oga

ugin

gpe

rfor

med

on10

0%of

the

part

saf

ter

part

sha

vele

ftth

est

atio

n.

5

Mod

erat

ely

high

Con

trol

sha

vea

good

chan

ceto

dete

ct.

XX

Err

orde

tect

ion

insu

bseq

uent

oper

atio

ns,

OR

gaug

ing

perf

orm

edon

setu

pan

dfir

st-p

iece

chec

k(f

orse

tup

caus

eson

ly).

4

Hig

hC

ontr

ols

have

ago

odch

ance

tode

tect

.X

XE

rror

dete

ctio

nin

-sta

tion,

orer

ror

dete

ctio

nin

subs

eque

ntop

erat

ions

bym

ultip

lela

yers

ofac

cept

ance

:su

pply

,se

lect

,in

stal

l,ve

rify

.C

anno

tac

cept

disc

repa

ntpa

rt.

3

Ver

yhi

ghC

ontr

ols

alm

ost

cert

ain

tode

tect

.X

XE

rror

dete

ctio

nin

-sta

tion

(aut

omat

icga

ugin

gw

ithau

tom

atic

stop

feat

ure)

.C

anno

tpa

ssdi

scre

pant

part

.2

Ver

yhi

ghC

ontr

ols

cert

ain

tode

tect

.X

Dis

crep

ant

part

sca

nnot

bem

ade

beca

use

item

has

been

erro

r-pr

oofe

dby

proc

ess/

prod

uct

desi

gn.

1

Insp

ectio

nTy

pes:

A.

Err

or-p

roof

ed;

B.

Gau

ging

;C

.M

anua

lIn

spec

tion.

Sour

ce:

Rep

rint

edw

ithpe

rmis

sion

from

the

FM

EA

Man

ual

(Dai

mle

rChr

ysle

r,Fo

rdM

otor

Com

pany

,G

ener

alM

otor

sSu

pplie

rQ

ualit

yR

equi

rem

ents

Task

Forc

e).

TA

BL

E14

.9E

xam

ple:

Pot

enti

alF

ailu

reM

ode

and

Eff

ects

Ana

lysi

s(P

roce

ssF

ME

A)

FME

ATy

pe(D

esig

nor

Proc

ess)

:Pr

oces

sPr

ojec

tN

ame

/Des

crip

tion:

Che

etah

/Cha

nge

surf

ace

finis

hof

part

Dat

e(O

rig.

):4

/14

Res

pons

ibili

ty:

Paul

aH

inke

lPr

epar

edB

y:Pa

ula

Hin

kel

Dat

e(R

ev.)

:6

/15

Cor

eTe

am:

Sam

Smith

,H

arry

Ada

ms,

Hilt

onD

ean,

Har

ryH

awki

ns,

Sue

Wat

kins

Dat

e(K

ey):

Des

ign

FME

A(I

tem

/Fu

nctio

n)Pr

oces

sFM

EA

(Fun

ctio

n/

Req

uire

men

ts)

Pote

ntia

lFa

ilure

Mod

ePo

tent

ial

Eff

ect(

s)of

Failu

re

S e v

C l a s s

Pote

ntia

lC

ause

(s)/

Mec

hani

sm(s

)of

Failu

re

O c c u rC

urre

ntC

ontr

ols

D e t e c

R P NR

ecom

men

ded

Act

ions

Res

pons

ibili

tyan

dTa

rget

Com

plet

ion

Dat

eA

ctio

nsTa

ken

S e v

O c c u r

D e t e c

R P N

Sold

erdi

ppin

gE

xces

sive

sold

er/

sold

erw

ire

prot

rusi

on

Shor

tto

shie

ldco

ver

9Fl

uxw

ire

term

inat

ion

610

0%in

spec

tion

316

2A

utom

atio

n/D

OE

/10

0%ch

kw

ithgo

/no

goga

ge

Sam

Smith

6/4

Don

e9

42

72

Inte

rloc

kba

seda

mag

eV

isua

lde

fect

s7

Lon

gso

lder

time

8A

utom

atic

sold

erto

ol3

168

Aut

omat

ion

/DO

E/

defin

evi

sual

crite

ria

Har

ryA

dam

s5

/15

Don

e7

42

56

7H

igh

tem

p8

Aut

omat

icso

lder

tool

/SP

C3

168

Aut

omat

ion

/DO

EH

ilton

Dea

n5

/15

Don

e7

42

56

Del

amin

atio

nof

inte

rloc

kba

seV

isua

lde

fect

s7

See

inte

rloc

kba

seda

mag

e8

Aut

omat

icso

lder

tool

/SP

C3

168

Aut

omat

ion

/DO

ESu

eW

atki

ns5

/15

Don

e7

42

56

7M

oist

ure

inin

terl

ock

base

5N

o7

245

Info

rmsu

pplie

rto

cont

rol

mol

ding

cond

.H

arry

Haw

kins

5/1

5D

one

72

798

Oxi

diza

tion

ofgo

lden

plat

ing

pins

Con

tact

prob

lem

/no

sign

al8

Not

bein

gcl

eane

din

time

7C

lean

in30

min

utes

afte

rso

lder

dip

528

0Im

prov

equ

ality

ofpl

atin

gde

fine

crite

ria

with

cust

omer

Sam

Smith

5/1

5D

one

82

580

Mar

king

Mar

king

perm

anen

cyte

stL

egib

lem

arki

ng/

cust

omer

unsa

tisfa

ctio

n

6 6 6

Mar

king

ink

Cur

ing

Smoo

thm

arki

ngsu

rfac

e

4 5 8

SPC

UV

ener

gyan

dSP

C

Non

e

2 3 6

48 90 288

Non

e

Non

e

Rou

ghsu

rfac

eSa

mSm

ith5

/15

Cha

nge

inte

rloc

kte

xtur

esu

rfac

e

63

610

8

Sour

ce:

Puls

e,a

Tech

nitr

olC

ompa

ny,

San

Die

go,

CA

(Jim

Fish

and

Mar

yM

cDon

ald)

.

TA

BL

E14

.10

Exa

mpl

e:P

oten

tial

Fai

lure

Mod

ean

dE

ffec

tsA

naly

sis

(Pro

cess

FM

EA

)FM

EA

Type

(Des

ign

orPr

oces

s):

Proj

ect

Nam

e/D

escr

iptio

n:B

usin

ess

oper

atio

nsof

Ato

Zim

port

sD

ate

(Ori

g.):

6/1

1

Res

pons

ibili

ty:

Prep

ared

By:

KC

Dat

e(R

ev.)

:7

/31

Cor

eTe

am:

KC

,JG

,L

MD

ate

(Key

):

Des

ign

FME

A(I

tem

/Fu

nctio

n)Pr

oces

sFM

EA

(Fun

ctio

n/

Req

uire

men

ts)

Pote

ntia

lFa

ilure

Mod

ePo

tent

ial

Eff

ect(

s)of

Failu

re

S e v

C l a s s

Pote

ntia

lC

ause

(s)/

Mec

hani

sm(s

)of

Failu

re

O c c u rC

urre

ntC

ontr

ols

D e t e c

R P NR

ecom

men

ded

Act

ions

Res

pons

ibili

tyan

dTa

rget

Com

plet

ion

Dat

eA

ctio

nsTa

ken

S e v

O c c u r

D e t e c

R P N

Bus

ines

sO

pera

tions

Shut

dow

nL

oss

ofin

com

e/

bank

rupt

cy9

Torn

ado

hits

loca

tion

3N

one

1027

0In

stal

lw

eath

erch

anne

lra

dio

inst

ore,

and

keep

ondu

ring

stor

eho

urs

JG7

/8In

stal

led

and

test

ed9

32

54

9L

awsu

itby

visi

tor

hurt

inst

ore

duri

ngvi

sit

3In

sura

nce

cove

rage

agai

nst

acci

dent

sin

stor

e

254

Non

e

9L

awsu

itby

visi

tor

owin

gto

faul

tym

erch

andi

se

5W

arni

ngla

bels

onm

erch

andi

se2

90N

one

9E

lect

rica

lfir

ebu

rns

dow

nst

ore

2Fi

reex

tingu

ishe

rsan

dsp

rink

lers

1018

0In

stal

lgr

ound

faul

tin

terr

upto

rs,a

ndov

erlo

ad/t

herm

alpr

otec

tion

onal

lhi

ghw

atta

gefix

ture

s

LM

6/2

8In

stal

led

GFI

san

dth

erm

alpr

otec

tion

92

118

9IR

Sau

dit

show

sm

isre

port

ing

offin

ance

s

5C

PAau

dits

acco

unts

atta

xtim

e

418

0C

hang

epr

oced

ure

toal

low

clos

ing

ofbo

oks

ever

y6

mon

ths,

and

CPA

toau

dit

the

sam

e

KC

7/1

5Pr

oced

ure

chan

ged,

acco

untin

gpe

rson

nel

and

CPA

info

rmed

92

236

9E

xces

sive

com

petit

ion

5A

gree

men

tw

ithpr

oper

tyow

ners

onlim

iting

num

ber

ofim

port

stor

es

290

Non

e

9L

oss

ofle

ase

10R

enta

lag

reem

ent

onm

onth

tom

onth

basi

s,au

tom

atic

ally

rene

ws

1090

0N

egot

iate

with

leas

ing

com

pany

toch

ange

leas

eag

reem

ent

toye

arly

KC

7/1

9Ta

lked

mat

ter

over

with

prop

erty

owne

rs,

obta

ined

verb

alas

sura

nce,

but

leas

est

ays

mon

thto

mon

th

910

1090

0

Ear

ning

sgr

owth

does

not

mee

tta

rget

s

Del

ayed

loan

repa

ymen

ts6

Sale

sst

aff

impo

lite

4Jo

bin

terv

iew

attim

eof

hiri

ng5

120

Inst

itute

sale

str

aini

ngof

new

hire

sfo

rha

lfda

yin

addi

tion

toex

istin

gtr

aini

ng.

Do

not

assi

gnto

floor

ifca

ndid

ate’

spe

rfor

man

cein

sale

str

aini

ngis

susp

ect

KC

8/2

Sale

str

aini

ngm

odul

ead

ded

toex

istin

gtr

aini

ngpa

ckag

e

62

224

6E

xces

sive

com

petit

ion

5A

gree

men

tw

ithpr

oper

tyow

ners

onlim

iting

num

ber

ofim

port

stor

es

260

Non

e

6Su

pplie

rde

lays

owin

gto

late

paym

ents

3N

one

1018

0C

ondu

ctFM

EA

onth

isca

use,

trea

ting

itas

afa

ilure

mod

eits

elf

6L

ocal

econ

omy

slow

s5

Non

e10

300

Mon

itor

area

grow

thth

ruqu

arte

rly

chec

ksw

ithth

elo

cal

Cha

mbe

rof

Com

mer

ce

JG7

/15

Obt

aine

dpo

pula

tion

grow

thfo

rci

ty,

and

inco

me

stat

istic

sfo

rqu

arte

ren

ding

Mar

ch31

st.

65

260

6St

ore

untid

y9

Em

ploy

ees

have

stan

ding

orde

rsto

atte

ndcu

stom

ers

first

;an

dup

keep

ofst

ore

seco

nd.

154

Non

e

TA

BL

E14

.10

(Con

tinue

d)

Des

ign

FME

A(I

tem

/Fu

nctio

n)Pr

oces

sFM

EA

(Fun

ctio

n/

Req

uire

men

ts)

Pote

ntia

lFa

ilure

Mod

ePo

tent

ial

Eff

ect(

s)of

Failu

re

S e v

C l a s s

Pote

ntia

lC

ause

(s)/

Mec

hani

sm(s

)of

Failu

re

O c c u rC

urre

ntC

ontr

ols

D e t e c

R P NR

ecom

men

ded

Act

ions

Res

pons

ibili

tyan

dTa

rget

Com

plet

ion

Dat

eA

ctio

nsTa

ken

S e v

O c c u r

D e t e c

R P N

6D

elay

edor

lost

ship

men

tsfr

omsu

pplie

r

7Fr

eigh

tfo

rwar

der

faxe

sbi

llof

ladi

ngw

hen

mer

chan

dise

islo

aded

onsh

ip/

air.

Nex

tco

ntac

tis

whe

ngo

ods

arri

veat

dest

inat

ion

1042

0R

equi

refr

eigh

tfo

rwar

der

toin

timat

est

atus

ofsh

ipm

ent

ever

y3

days

KC

7/2

8G

otag

reem

ent

with

frei

ght

forw

arde

r;ad

ditio

nal

char

geof

2.5%

onfr

eigh

tag

reed

upon

67

312

6

6D

efec

tive

mer

chan

dise

4In

spec

tion

prio

rto

putti

ngm

erch

andi

seon

shel

ves

248

Non

e

6T

heft

ofca

shby

empl

oyee

sfr

omca

shre

gist

ers

7L

ogs

ofem

ploy

eena

mes

man

agin

gca

sh,

byda

tean

dtim

e

1042

0Su

perv

isor

sto

star

tac

coun

ting

for

cash

with

empl

oyee

sw

hen

they

star

tw

ork

and

atev

ery

chan

geov

er.

KC

8/3

Proc

edur

espu

tin

plac

eto

acco

mpl

ish

cash

man

agem

ent

asre

com

men

ded

62

224

6T

heft

ofm

erch

andi

se5

Stor

eat

tend

ants

tom

onito

rcu

stom

ers

whe

nfe

asib

le

824

0In

stal

lm

agne

ticth

eft

prev

entio

nta

gson

mer

chan

dise

and

dete

ctor

sat

stor

een

tran

ce

LM

8/1

5C

ompl

eted

onite

ms

with

ticke

tpr

ices

over

$20;

rest

will

beco

mpl

eted

by8

/15

65

130

6W

rong

mer

chan

dise

lead

ing

tosl

ower

inve

ntor

ytu

rns

5V

isit

who

lesa

lem

arke

tstw

ice

/yea

rto

keep

upw

ithcu

rren

ttr

ends

390

Non

e

6A

ccou

ntin

ger

rors

5B

ooks

audi

ted

byC

PAat

tax

time

412

0N

one

Sour

ce:

Rai

Cho

wdh

ary.

fmea.pdf

Documents

Transcript of fmea.pdf