FMEA: A Problem Prevention Tool

 Dr. Edgardo J. Escalante.

 (*) Missile Defense Agency

 Dr. Edgardo J. Escalante.

A story…

We have designed a 5-floor

commercial aircraft that has
-tennis courts
-restaurants
-boutiques

The only problem is that we haven’t

verified if it can fly!!!
 Dr. Edgardo J. Escalante.
 Quality “Control”

RELIABILITY
-fmea

 Dr. Edgardo J. Escalante.

 The production
elements
1.- Men
2.- Materials
3.- Machines and equipment
4.- Methods
5.- Information

NKS/Factory Magazine (1988)

 Dr. Edgardo J. Escalante.

An important missing element.
What is it?

Information Materials Machines

Methods Men

 Dr. Edgardo J. Escalante.

…right!

Information Materials Machines

CONTROL

Methods Men

 Dr. Edgardo J. Escalante.

 What does Control
mean?

To take actions to keep

a situation in a certain
desired state

 Dr. Edgardo J. Escalante.

Definition of FMEA Failure Mode and
Effects Analysis
A systemized group of activities designed to

1. RECOGNIZE and EVALUATE the POTENTIAL

FAILURE of a PRODUCT/PROCESS and its EFFECTS

2. IDENTIFY ACTIONS WHICH COULD ELIMINATE

or REDUCE the CHANCE of POTENTIAL FAILURE
OCURRING

3. DOCUMENT the PROCESS

(D,F,GM. FMEA, 2001)

 Dr. Edgardo J. Escalante.
Two types of FMEA

-Design FMEA or DFMEA

-Process FMEA or PFMEA

 Dr. Edgardo J. Escalante.

Characteristics of FMEA
-It looks to minimize the probability of a failure, or to
minimize its effects

-Should be initiated before or at design concept

finalization (DFMEA), or before production starts
(PFMEA)

-It is an endless iterative process

-It is a way to document processes and designs

 Dr. Edgardo J. Escalante.

 The DFMEA evaluates what can go wrong
with the product during its use and
during its manufacturing as a consequence
of design weaknesses
(Aldridge, Taylor 1991)

The PFMEA focuses on the reasons of

potential failures during manufacturing as a result
of not complying with the original design or
failing to meet the design’s specifications
(Aldridge, Taylor 1991)

 Dr. Edgardo J. Escalante.

History of FMEA

Originally used by NASA during the

60s. Ford Motor Co. started using it
during the 70s

(Gilchrist, 1993)

 Dr. Edgardo J. Escalante.

 FMEA benefits

-Reduction of internal cost due to re-work for not

doing things right the first time

-Reduction of the number of complaints and warranty

costs

-Increase of customer satisfaction

-Confidence that a company’s products are based in

reliable and robust production methods
(Aldridge, Taylor 1991)
 Dr. Edgardo J. Escalante.
 Potential
Failure Mode and Effects Analysis

S C O D R Actions Results
e l c e P
Item Potential Potential v a Potential Causes/ c Current t Responsibility
Failure Effect(s) of s Mechanisms(s) u Controls e N Recommended & Target S O D R
s Failure r c Action(s) Completion Date Actions e c e P
Function Mode Failure Prevention/Detection Taken v c t N

What can be done?

What How
- Design changes
are the bad
Effect(s)? is it? What did they
- Process changes do and what
How
What are the are the
often
nctions, features - Special controls outcomes
does
r requirements? it
happe - Changes to standards,
n? procedures,
What can go or guides
What are
wrong? the Cause(s)?
- No Function
How
- Partial/ Over/ good is
Degraded this
Function method
at Who is going
- Intermittent
How can thisdetecting to do it and
Function be prevented it? when?
- Unintended and detected?
Function

Summary of DFMEA/PFMEA (D,F,GM. FMEA, 2001)

 Dr. Edgardo J. Escalante.

 Definitions

Failure Mode: is the way in which the failure is

manifested

Failure effect: is the consequence of the failure

Failure cause: is what induces the failure

 Dr. Edgardo J. Escalante.

 Example (Ford, 1991; D,F,GM. FMEA,2001) PFMEA
C
Process/ Op. Potential Potential S l Potential causes
failure mode e a /
description failure effects v
and purpose s mechanism
of failures

Manual applicationInsufficiente
of Deteriorated life of 7
wax Manually inserted
wax inside door coverage over spray head not
door leading to:
specified surface inserted far enough

Unsatisfactory appea-
To cover inner door rance due to rust
lower surfaces at min.
wax thickness to through paint over
retard corrosion time

 Dr. Edgardo J. Escalante.

 Example (Ford, 1991; D,F,GM. FMEA,2001) PFMEA (cont)
O D R Recommended
Responsibility
c Current e P actions & target
u process t N completion
r control e date
r c
Visual
8 inspection 5 280
Add positive depth stop R. López
every hour. Verify thick- to sprayer 0X 10 15
ness and coverage with
gauge once per shift.(D).
Test spray pattern at Automatic spraying R. López
start-up and after idle
periods. (P). 0X 12 15

 Dr. Edgardo J. Escalante.

 Example (Ford, 1991; D,F,GM. FMEA,2001) PFMEA
ACTION RESULTS
Actions S O D R
taken e c e P
v u t N
r c
Stop added, sprayer checked on line
7 2 5 70

Rejected due to complexity of

different doors on same line

 Dr. Edgardo J. Escalante.

General Recommendations (Palady 1995)

1. To determine the effects it’s important to capture the final

customer experiences. Try to experiment what the customer
would feel. Otherwise severity may be underestimated.

2. To define severity, take into consideration safety issues as well

as costs.

3. The recommendations must be justified based on its cost-benefit

relationship, and must have a high degree of permanence.

 Dr. Edgardo J. Escalante.

 References
Aldridge J., Taylor J.(1991) “The Application of Failure Mode and
Effects Analysis at an Automotive Components Manufacturer”.
International Journal of Quality and Reliability Management, Vol.
8 No. 3.

Chrysler Corp., Ford Motor Co., General Motors Corp.(1995).

Potential Failure Mode and Effects Analysis. 2d. ed. A.I.A.G.

DaimlerChrysler Corp., Ford Motor Co., General Motors Corp.

(2001). Potential Failure Mode and Effects Analysis. 3rd. ed.
A.I.A.G.

Gilchrist W. (1993) “Modelling Failure Modes and Effects

Analysis”. International Journal of Quality and Reliability
Management, Vol. 10 No. 5.

NKS/Factory Magazine (1988). Poka Yoke. Improving Quality by

Preventing Defects. Productivity.
Palady
 P. (1995) Failure Modes And Effects Analysis. PT
Dr. Edgardo J. Escalante.
 APPENDIX

 Dr. Edgardo J. Escalante.

 DFMEA SEVERITY EVALUATION CRITERIA
EFFECT CRITERIA: Severity of Effect RANK

Hazardous- Very high severity ranking when a potential failure mode affects safe vehicle
without operation and/or involves noncompliance with government regulation without 10
warning warning

Hazardous- Very high severity ranking when a potential failure mode affects safe vehicle
with operation and/or involves noncompliance with government regulation with 9
warning warning

Very High Vehicle/item inoperable (loss of primary function). 8

Vehicle/item operable but at a reduced level of performance. Customer very
High 7
dissatisfied.

Vehicle/item operable but Comfort/Convenience item(s) inoperable. Customer

Moderate 6
dissatisfied.

Vehicle/item operable but Comfort/Convenience item(s) operable at a reduced

(D,F,GM. FMEA, 2001)

Low 5
level of performance. Customer somewhat dissatisfied.

Fit & Finish/Squeak & Rattle item does not conform. Defect noticed by most
Very Low 4
customers (greater than 75%).

Fit & Finish/Squeak & Rattle item does not conform. Defect noticed by 50% of
Minor 3
customers.

Fit & Finish/Squeak & Rattle item does not conform. Defect noticed by
Very Minor 2
discriminating customers (less than 25%).

None No discernible effect. 1

 Dr. Edgardo J. Escalante.

 Probability of Likely Failure Rates Over Design Ranking
Failure Life
DESIGN OCCURRENCE EVALUATION Very High:
Persistent
≥ 100 per thousand vehicles/items 10

failures
50 per thousand vehicles/items 9

High: 20 per thousand vehicles/items 8

Frequent
failures
10 per thousand vehicles/items 7

Moderate: 5 per thousand vehicles/items 6

Occasional
failures
2 per thousand vehicles/items 5

(D,F,GM. FMEA, 2001)

1 per thousand vehicles/items 4

Low: Relatively 0.5 per thousand vehicles/items 3

few failures
0.1 per thousand vehicles/items 2

Remote: ≤ 0.01 per thousand vehicles/items 1

Failure is
unlikely
 Dr. Edgardo J. Escalante.
 Detection Criteria: Likelihood of DETECTION by Design Ranking
Control
Absolute Design Control will not and/or can not detect a potential 10
Uncertainty cause/mechanism and subsequent failure mode; or there
is no Design Control.
Very Remote Very remote chance the Design Control will detect a 9
potential cause/mechanism and subsequent failure
mode.
Remote Remote chance the Design Control will detect a 8
potential cause/mechanism and subsequent failure
mode.
Very Low Very low chance the Design Control will detect a 7
potential cause/mechanism and subsequent failure
mode.
Low Low chance the Design Control will detect a potential 6
cause/mechanism and subsequent failure mode.

(D,F,GM. FMEA, 2001)

 Dr. Edgardo J. Escalante.
Moderate Moderate chance the Design Control will detect a 5
potential cause/mechanism and subsequent failure
mode.
Moderately Moderately high chance the Design Control will detect 4
High a potential cause/mechanism and subsequent failure
mode.
High High chance the Design Control will detect a potential 3
cause/mechanism and subsequent failure mode.
Very High Very high chance the Design Control will detect a 2
potential cause/mechanism and subsequent failure
mode.
Almost Design Control will almost certainly detect a potential 1
Certain cause/mechanism and subsequent failure mode.

 Dr. Edgardo J. Escalante.

 rd edition this table is presented in two columns but the information is
Effect PROCESS Ranking
------------------------------------------------------
Criteria: SEVERITY of Effect
Very high severity ranking when a potential
Hazardous- failure mode affects safe operation and/or involves 10
without noncompliance with regulations without warning.
warning Or may endanger operator without warning.

r effect and manufacturing/assembly effect).

Very high severity ranking when a potential
Hazardous- failure mode affects safe operation and/or involves 9
with warning noncompliance with regulations with warning.
Or may endanger operator with warning.

Product/item inoperable, with loss of primary

Very high function. Or 100% of product may have to be 8

(D,F,GM. FMEA, 2001)

scrapped or repair time >1h.

Product/item operable, but at reduced level of

High performance. Customer dissatisfied. Or product 7
may have to be sorted and less than 100%
scrapped. Repair time between 0.5 and 1h.
Product/item operable, but comfort/convenience
Moderate inoperable. Customer dissatisfied. Or a portion 6
(less than 100%) of the product may have to be
scrapped with no sorting. Repair time less than 0.5h
 Dr. Edgardo J. Escalante.
 Product/item operable, but comfort/convenience
Low operable at a reduced level. Or than 100% of the 5
product may have to be reworked with no sorting.
Repairing off-line.

Fit and finish/squeak and rattle item does not

Very Low conform. Defect noticed by most customers 4
(greater than 75%). Or product may have to be
sorted with no scrap, and a portion (less than 100%
reworked.

Fit and finish/squeak and rattle item does not

Minor conform. Defect noticed by 50% of the customers. 3
Or a portion (less than 100%) of the product may
have to be reworked with no scrap. On-line but
out-of-station.
Fit and finish/squeak and rattle item does not
Very Minor conform. Defect noticed by discriminating 2
customers (less than 25%). Or a portion (less than
100%) of the product may have to be reworked
with no scrap. On-line but in-station.

None No discernible effect. Or slight inconvenience to 1

operation or operator, or no effect.

(D,F,GM. FMEA, 2001)

 Dr. Edgardo J. Escalante.
 Probability of Likely Failure Rates Ranking
Failure
PROCESS OCCURRENCE RANKING Very High: ≥100 per thousand pieces 10
Persistent
failures
50 per thousand pieces 9

High: Frequent 20 per thousand pieces 8

failures
10 per thousand pieces 7

Moderate: 5 per thousand pieces 6

Occasional
failures
2 per thousand pieces 5

(D,F,GM. FMEA, 2001)

1 per thousand pieces 4

Low: 0.5 per thousand pieces 3

Relatively few
failures
0.1 per thousand pieces 2
Remote: ≤ 0.01 per thousand pieces 1
Failure is
unlikely

 Dr. Edgardo J. Escalante.

 DETECTION CRITERIA FOR PFMEA
(D,F,GM. FMEA, 2001)
Criteria A B C Suggested Range of Detection Methods Ranking
Detection
Almost Absolute certainty Cannot detect or is not checked. 10
Impossible of non-detection.
Very Remote Controls will Control is achieved with indirect or random checks only. 9
probably not
detect.
Remote Controls have
poor chance of
Control is achieved with visual inspection only. 8
detection.
Very Low Controls have Control is achieved with double visual inspection only. 7
poor chance of
detection.
Low Controls may Control is achieved with charting methods , such as SPC 6
detect. {Statistical Process Control}.
Moderate Controls may Control is based on variable gauging after parts have left 5
detect. the station, OR Go/No Go gauging performed on 100% of
the parts after parts have left the station.
Moderately Controls have a Error detection in subsequent operations, OR gauging 4
High good chance to performed on setup and first- piece check.
detect.
High Controls have a Error detection in- station, OR error detection in subsequent 3
good chance to operat ions by multiple layers of acceptance: supply, select,
detect. install, verify. Can not accept discrepant part.
Very High Controls almost
certain to detect.
Error detection in- station (automatic gauging with automatic 2
stop feature). Can not pass discrepant part.
Almost certain Controls certain to Discrepant parts can not be made because item has been 1
detect. error proofed by process/product design.

tion types: A=Mistake-proofed, B=Gauging, C=Manual inspe

 Dr. Edgardo J. Escalante.