DFMEA Guidelines(500 fortune)

Design Failure Mode and Effects Analysis Guidelines
























Revision 6 - May 24, 2006
Introduction
This Intier Automotive Seating DFMEA Manual is intended to act as a supplemental guideline to the official AIAG manual. It does not replace or supercede the AIAG manual.
The objective of this manual is to make clear the concepts, methodologies, and approaches of the AIAG manual by providing seating-specific information and applications.

Any audit disputes regarding DFMEA practices and methodologies must be resolved in favor of the latest AIAG FMEA manual.


Customer Defined

The DFMEA must consider the needs of all potential customers. The most common are:
 The Consumer (end-customer)
 The OEM
 The JIT plant / manufacturing
 Service departments

Completing the DFMEA Document – Header Area

1) FMEA Number

The FMEA document number that will be used for tracking. Use the Intier serialized numbering system described below:
(program suffix)-(seat row)-(activity acronym)-(serial number)
Example: C7CX-DRV-DFMEA-001

2) System, Subsystem, or Component Name and Number

System DFMEA Scope
The System is the Tier 1 assembly. In the case of Intier Automotive Seating, this applies specifically to a Seat Complete and is broken out by row. For example: 2nd row quad seat, 3rd row SO/50 split bench seat, or powered driver's seat. This DFMEA is to be completed by the seating team.

Subsystem DFMEA Scope
Subsystems are Tier 2 assemblies such as recliners, power adjusters, lumbar systems, head-restraint sleeves, and trim covers. They are always end-items that are supplied to the Tier 1 JIT plant. The Tier 2 design-responsible supplier is responsible for completing these DFMEAs.

Component DFMEA Scope
Component level DFMEAs are generally completed for significant sub-assemblies within an end-item assembly. Recliner disks and stamped assemblies are typical examples. The design-responsible component supplier is required to complete these DFMEAs.


3) Design Responsibility
See AIAG FMEA Manual 3rd Edition.

4) Prepared By
See AIAG FMEA Manual 3rd Edition.

5) Model Year(s) / Program(s)
See AIAG FMEA Manual 3rd Edition.

6) Key Date
Enter the initial DFMEA due date, which should not exceed the scheduled production design release date.

7) FMEA Date
Enter the date the original DFMEA was completed and the latest revision date.

8) Core Team
See AIAG FMEA Manual 3rd Edition

Completing the DFMEA Document – Main Body

9) Item / Function
The Intier DFMEA approach considers Functions, not Items. Hence, enter all the functions expected from the seat system, or end-item / component, in terms of comfort, functionality, quietness, ergonomics, efforts, packaging, etc. Each function should be listed separately.

For Tier 1 seat complete DFMEAs, use the prescribed Intier DFMEA template to start the analysis. Typical seat complete functions could include, but are not limited to:

 Provide acceptable comfort
 Meet applicable safety and structural requirements
 Provide acceptable ergonomics of handles, straps, switches, etc. Provide acceptable BSR performance in vibration environments Meet customer serviceability requirements
 Provide easy assembly at JIT plant
 Comply with customer mass requirements
 Provide acceptable appearance
 Provide electrical connection to vehicle

End-item (Tier 2) and Component level (Tier 3) DFMEAs must comprehend all relevant functions from the seat complete DFMEA and may contain additional functions that are more specific to the end-item or component. Trim and Foam DFMEAs are treated as End-Item DFMEAs.


10) Potential Failure Mode

The Potential Failure Mode should describe how the end-item fails. For Seat Complete DFMEAs, typical failure modes could be, but are not limited to:

 Back-frame main tube fractures or yields under specified load
 Recliner disengages under specified load
 Trim wrinkles
 Foam sets or tears
 Fasteners come loose
 Heating element overheats
 Component interferes with adjacent parts
 Adjuster binds
 Material corrodes
 Frame does not support foam

End-item and Component level DFMEAs have similarly structured failure modes

Potential Failure Modes should be described in physical or technical terms, not as a symptom necessarily noticeable by the customer.

11) Potential Effect(s) of Failure

The Potential Effect(s) of Failure describe, in non-technical terms, how the customer could be affected. Consider the potential effects on all customers. Typical failure effects could be, but are not limited to:

 Potential injury to occupant with warning
 Potential injury to occupant without warning
 Non-compliance with federal specifications
 Poor appearance
 Operation impaired
 Generates noise
 Seat is unstable during driving conditions
 Unable to assemble at JIT plant
 Difficult to service
 Hazardous to operator / occupant


12) Severity

The Severity Table below describes what each ranking level represents and includes seating-specific examples.

Table 1. DFMEA Severity Evaluation Criteria
Effect Criteria (Severity of Effect) Ranking
Hazardous, without warning Very high severity ranking when a potential failure mode affects safe vehicle operation and/or involves noncompliance with government regulation without warning. Examples: Welds breaking, latches disengaging, fasteners breaking. 10
Hazardous, with warning Very high severity ranking when a potential failure mode affects safe vehicle operation and/or involves noncompliance with government regulation with warning. Examples: Lazy latches that indicate they will disengage in a crash, tracks do not engage, heating element burns through fabric, any serious hazards pre-alerted with warning labels. If unsure whether to pick a severity of 9 or 10, choose 10. 9
Very High Seat system or end-item inoperable (loss of function). Does not endanger occupant. Examples: Power tracks seize, latches bind, parts do not go together during assembly, item cannot be serviced. 8
High Seat system or end-item operable but at a reduced level of performance. Customer very dissatisfied. 7
Moderate Seat system or end-item operable but comfort/convenience item(s) inoperable. Customer dissatisfied. Examples: Cupholder jams, armrest binds, lumbar mechanism locks up, heating element does not work. 6
Low Seat system or end-item operable but comfort/convenience item(s) operable at a reduced level of performance. Customer somewhat dissatisfied. Examples: Armrest has high efforts, lumbar mechanism doesn't operate to specified
range, cup holder doesn't retain drink containers well, heating element emits inconsistent heat. 5
Very Low Fit / finish or BSR item. Defect noticed by most customers (greater than 75%). Examples: bad wrinkles on trim, seat squeaks loudly, dislodged shield, exposed screw-head 4
Minor Fit / finish or BSR item. Defect noticed by 50% of customers. Examples: wrinkles on trim, large gaps between foam and plastic, rattling in seat during driving conditions, misalignment between split seat-backs, mismatched hues / colors. 3
Very Minor Fit / finish or BSR item. Defect noticed by discriminating customers (less than 25%). Example: slight squeak during seat operation, minor wrinkles on trim, gaps between parts in discrete areas, minor color mismatches. 2
None No discernible effect 1

13) Classification

Any Potential Failure Modes that have effects ranked as a severity 9 or 10 must be flagged in the classification column as "CC" (Critical Characteristic). Such characteristics require additional design or process controls.
Customer specific symbols may be used in lieu of "CC" but such practice is not recommended.

Whenever an item in the DFMEA is classified as a CC, there must be a Recommended Action put in place to complete one of the following activities:

1) If Intier is design responsible for the component then state: “Cascade CCs to appropriate engineering drawing(s)”
2) If Intier is not design responsible for the component then state: "Cascade CCs to appropriate supplier DFMEA"

All CCs in the DFMEA must translate into engineering drawings or supplier DFMEAs.

The target date for ensuring CCs cascade into engineering drawings or supplier DFMEAs must not exceed the DFMEA Key Date. Actual completion date should be as early as possible in the program.

In the unusual event that the program team determines that it is not necessary to cascade a CC identified in the DFMEA there must be written documentation of that decision in the Actions Taken column.

14) Potential Cause(s) / Mechanisms of Failure

All Causes should begin with a negative connotation, indicating a misjudgment in the design of the system, subsystem or component. Usual verbiage to start off a Cause include, but are not limited to:

 Improper
 Inadequate
 Incorrect
 Excessive
 Insufficient

Typical Causes may include, but are not limited to:

 Improper material specified
 Improper structural weld call-outs
 Excessive bolster height
 Improper torque specified
 Improper material section along load path Improper friction material specified
 Improper tolerance call-out
 Inadequate trim retention specified
 Insufficient clearance allowed
 Excessive spring rate


15) Occurrence

Table 2. DFMEA Severity Evaluation Criteria
Probability of Failure Description and Estimated Failure Rates Ranking
Very High: Persistent Failures Primarily from new designs / ideas that are untested. undocumented and unknown to the team.

100 failures per thousand seats produced / tested. 10
Primarily from new designs / ideas that are untested. undocumented and unknown to the team.
50 failures per thousand seats produced / tested. 9
High: Frequent Failures Newer designs/ideas that have been developed in other arenas but are untested and have yet to be fully understood in the seating context.
20 failures per thousand seats produced / tested. 8
Newer designs / ideas that have been developed in other arenas but are untested and have yet to be fully understood in the seating context.
10 failures per thousand seats produced / tested. 7
Moderate: Occasional Failures Known seating design/idea that has yet to be fully perfected. but continuous improvement is underway on other programs to improve performance.
5 failures per thousand seats produced / tested. 6
Known seating design/idea that has yet to be fully perfected. but continuous improvement is underway on other programs to improve performance.
2 failures per thousand seats produced / tested. 5
Known seating design/idea that has yet to be fully perfected. but continuous improvement is underway on other programs to improve performance.
1 failure per thousand seats produced / tested. 4
Low: Relatively Few Failures Adaptation of proven seating design/idea being used on a different platform. Issues exist but are predictable and can be continuously improved.
0.5 failure per thousand seats produced / tested. 3
Adaptation of proven seating design/idea being used on a different platform. Issues exist but are predictable and can be continuously improved.
0.1 failure per thousand seats produced / tested. 2
Remote: Unlikely Failure Reliable complete carryover design / idea or designs that have been in service reliably for over a year. All major issues have been closed and most continuous improvements are not needed or very minor.
< 0.010 failure per thousand seats produced/tested. 1

16) Current Design Controls

Current Design Controls are engineering studies, theoretical or physical, that have been, or are being, used with the same or similar product designs (controls with "proven value"). There are two types of design controls to consider:

Prevention - These prevent the cause / mechanism of the failure mode from occurring, or reduce the rate of occurrence. These are usually theoretical, analytical or forecasting type measures that occur early in product development to prevent issues before they occur on physical parts. Examples of Prevention Controls: mathematical calculations, FEA analysis, packaging studies, supplier DFMEAs, and design reviews.

If no Prevention Design Controls exist, state "No current prevention design control". Absence of a prevention design control typically results in a higher Occurrence score.


Detection - These detect the cause / mechanism of the failure mode by physical methods before the product is release for production. Examples of Detection Controls: customer approved DVP&R testing, prototype builds, fleet reviews, ride and drive's, acid etch tests, torque strip-out tests, pressure mapping, cut & weld evaluation, and jury evaluations.

If no Detection Design Controls exist state "No current detection design control". Lack of a Detection Control will yield a 10 for Detection ranking. In actual practice it is rare to have no detection design control.

For safety related items (severities 9 and 10) actual federal or safety tests are listed under Detection Controls. Example: "Customer approved DVP&R - FMVSS 225 testing".

Prevention and Detection Design Controls are tasks the design engineering team is committing to completing before production release. Therefore, the design controls should be tracked in an open issues matrix format (MIDS or similar system).

17) Detection

Similar to AIAG FMEA Manual 3rd Edition.

Table 3. Suggested values for Detection Controls
Design Control Ranking
Federal safety testing 1-2
Compliance Manual 1-4
Dynamic sled testing (due care) 2-4
Life cycle, fatigue, durability testing 2-5
Non-safety static loading / deflection testing 2-4
BSR testing 4-8
Comfort evaluations 3-6
Appearance / craftsmanship evaluations 4-8
Low volume build reviews 4-8
Jury evaluation clinics 3-8
Abuse clinic 3-6
Manufacturing review 2-6
DOEs 2-4


18) Risk Priority Number (RPN)

The Risk Priority Number is the product of the Severity, Occurrence, and Detection
rankings. This number may be used to rank the concerns within the design.
One purpose of the DFMEA is to highlight higher risk areas of the design. As such, the team must choose a strategy for addressing these higher risk areas. The chosen strategy must be entered in the DFMEA header in the appropriate location. The activities undertaken to reduce the design risk are listed in Recommended Actions. The strategy may change over the life of the design.

Typical Strategies:

&#61656; Address the items with the highest RPN ranking
&#61656; Address the items with the highest Severity ranking
&#61656; Address the items with the highest Criticality ranking
(Criticality is the product of the Severity and Occurrence rankings)

For any of the above approaches...
Select a specific number of items to address at one time (e.g. Top 20) or
Select a specific percentage of items to address at one time (e.g. Top 10%)

Do not use a specific RPN cut-off value as a trigger

19) Recommended Action(s)

Recommended Actions are a permanent list of actions that document the history of additional tasks, beyond the prevention and detection controls, that the team is committed to completing. For example:

&#61656; Cascade CCs to appropriate engineering drawing(s)
&#61656; Cascade CCs to appropriate supplier DFMEA
&#61656; Revise test plan or test method(s)
&#61656; Develop new test(s)
&#61656; Investigate use of new material(s)
&#61656; Revise material specifications

If engineering assessment leads to no recommended actions for a specific failure mode / cause / control combination then the team can state "None" for recommended actions in accordance with AIAG guidelines.

20) Responsibility for the Recommended Action(s)

See AIAG FMEA Manual 3rd Edition.

21) Action(s) Taken

Record applicable PCN changes, results of completed Recommended Actions, and general notes regarding changes to the DFMEA line item.

Use the following format:

Date of Update: Tracking Number - Brief Description of Action/Change. How RPN was affected.
Example: 04-Apr-06: PCN 12345 - Changed SIAB bracket material - Occurrence reduced from 4 to 3.



Every Action Taken must state the effect on Severity, Occurrence, and / or Detection rankings. Examples: "Occurrence increased from 3 to 4", "Detection reduced from 5 to 3". If the rankings did not change write "Rankings not affected".

All Actions Taken must be documented and that documentation must be available for viewing upon request. Only documented actions should be in a DFMEA - verbal or non- reported activities cannot be cited as Actions Taken. For tracking numbers use either the relevant document number (e.g. PCN #, DWO #, MIDS #) or the Intier serialized numbering system described earlier in this document.
Note that for a given PCN multiple causes / failure modes may be affected. Thus, the description of the actions taken may be required to be added to the DFMEA in multiple locations.

22) Action Results

After the preventative / corrective action has been identified, estimate and record the resulting severity, occurrence, and detection rankings. Calculate and record the resulting RPN.
All revised ratings should be reviewed. If further action is considered, repeat the analysis. The focus should always be on continuous improvement.
If an action taken does not alter the original RPN, the team is still required to populate the Action Results column with the previous Severity, Occurrence, Detection and RPN values. These values should remain the same as the original rankings. Assure that the verbiage of the action taken states "Rankings not affected".

Frequently Asked Questions

This section is intended to address situations that frequently occur during DFMEA development and subsequent updates.

Q) Do I change my original Severity, Occurrence, and Detection rankings after initial
design release?
A) No. Once you have released the design and entered PCN mode, the original content of the DFMEA will not change. All updates to the DFMEA will be entered in the Action Results section. Prior to initial design release the Severity, Occurrence, and Detection may change as the design evolves.

Q) Does every PCN affect the DFMEA?
A) No. Some PCNs will not effect the DFMEA (e.g., color changes, drawing note changes). If this is the case, include "No change to DFMEA" in the Description of Revision section in the Revision Log

Q) If I write a PCN that affects the DFMEA, do I need to enter a recommended action?
A) No. Just enter the Action Taken and Action Results per 21 and 22 above.

Q) Does the RPN need to change for every PCN that affects the DFMEA?
A) No. The Severity, Occurrence, and Detection should be re-evaluated after every
design change. If no change is required state "Rankings not affected" per 22 above.

Q) Do I need to carry the RPN to the Action Results section if I do not have any
Actions taken?
A) No. If there have not been any Actions Taken then the original Severity, Occurrence, and Detection are still current.

Q) If I do not have any Recommended Actions do I need to write "None" in the Responsibility and Actions Taken sections?
A) No. If you have "None" in the Recommended Actions section leaving the
Responsibility and Actions Taken sections blank is acceptable.

Q) Is it acceptable to have no Prevention or Detection design control?
A) Yes. Fortunately, this situation is very rare. If this is the case then state: "No current design control" per 16 above.

Q) Do I need to have a cross-functional team at every meeting?
A) No. However, you need to show that all members of the core team have had
sufficient input into the DFMEA. For example, you can have a special DFMEA review with the warranty team members, etc.

Q) Do I need to have a cross-functional team review the DFMEA for every PCN?
A) Yes. By nature of the PCN process the DFMEA is included in the PCN documentation. The reviewers of the PCN are approving the DFMEA changes when they approve the PCN.

Q) When may the Potential Failure Mode be attributed to the top level assembly
instead of to an end-item?
A) The Potential Failure Mode may be attributed to the top level assembly when either of the following conditions exist:

[]The Failure Mode cannot be attributed to any single end-item[/]

Example: Function = Seat must be 80% recyclable by weight

[]The Failure Mode can be attributed to every end-item in the assembly[/]

Example: Function = Seat must not create offensive odors

Q) I am releasing a PCN that only contains updates for parts or assemblies designed by the customer. How do I update my DFMEA?
A) No update to the Intier DFMEA is required. See WI-7085 for requirements related to the PCN process.
Revision History

Rev Revision Date Revision Description Author
4 07-July-05 Made minor correction to the RPN cutoff value on page 9 C. Ngiau
5 10-March-06 Removed requirement to automatically address SCs in the DFMEA. Revised instructions on how to address CCs.
Removed requirement to automatically address RPNs with a value of 120 or more and changed the statement required in the Recommended Actions from “Per Intier standards listed in F-030 no recommended actions are required for items with an RPN below 120” to “None” for cases where no action is required.
Eliminated unnecessary "Special Attention" definition.
Eliminated instructions and attachment for Subjective Evaluation Plan & Report (obsolete).
Revised the DFMEA Guidelines to include recommended best practices and added Frequently Asked Questions section.
Revisions to address the obsolescence of Class I PCNs. DJ Gray
C. Hofer
6 24-May-07 Converted document from Microsoft Excel to Microsoft Word.
Added Revision History page.
Removed statement regarding linkage between Functions and Customer Requirements.
Add FAQ pertaining to customer-designed parts C. Hofer
DJ Gray