Aspects in the FMEA Grid

STRUCTURE ANALYSIS (STEP 2)

Structure Analysis aims to identify and break down the manufacturing process and represent the process flow as it physically exists. It provides the basis for the Function Analysis (Step 3)

Process Item

The Process Item of the PFMEA is at the top level of the PFMEA. In AVIX, this is always an assembly Line.

Process Step

The process step is a workstation or a manufacturing operation. It is also considered to be the focal element of the PFMEA. In AVIX, the process step and the focus element are separated. The process step is the Workstation, and the focus element is the Process Task.

Process Work Element

The Process Work Element is the lowest level of the process structure tree. Each work element is a potential cause that can impact a process step or process task and should be considered separately during the FMEA. Each work element corresponds to a category commonly used in Ishikawa causality diagrams and populated by lean manufacturing as 4M or 5M.

Example categories:

Machine
Man
Material (Indirect)
Environment (Milieu)
Method

Function analysis (step 3)

The process function analysis ensures that the functions and requirements of the product/process are taken care of. It provides the bases of the failure analysis (Step 4)

About Functions

A function describes what something is intended to do. A process item or process step can have multiple functions.
Before the team can perform the function analysis, they must know the product and process functions/requirements. This information can be gathered before or during the analysis.
The description of a process function should be clear and, if possible, phrased with a verb and a noun like "Positioning rear window" or "Fasten seat".

The function of the Process Item

A high-level function that references the Process Item in the Structure Analysis, it can take into account functions such as:

Internal function
External function
Customer and/or end-user function

The negation of the process item functions will be the Failure effects in the Failure analysis (step 4)

Example: "Assemble front seat"

The functions of a Process Step

The Functions of a Process Step describes the resulting product features produced at the workstation or the process task.

The negation of a process step function will be a Failure mode in the Failure analysis (step 4)

Example: "Fasten the seat with screws"

The functions of a Work Element

The functions of a Process Work Element reflect its contribution to the Process Step (Process Task) to create the process/product features.

The negation of a work element function will be a Failure cause in the Failure analysis (step 4)

Example: "Enter the screws into the threaded holes"
or
Example: "Tighten the screws with the right torque"

Failure Analysis (step 4)

Process-step failure causes, modes, and effects are identified in the failure analysis, and their relationship, the failure chain, is established. It provides the bases of the risk analysis (Step 5)

Failure analysis is performed for each element/step in the process description (Structure Analysis/Step 2 and Function Analysis/Step 3).

Failures

Failures of a process step are deduced from product and process characteristics. Examples include:

non-conformities
inconsistently or partially executed tasks
unintentional activity
unnecessary activity

Failure Effects (FE)

~~-------------------------------------------------------------------------------~~

Failure Effects are related to functions of the process item (System, Subsystem, Part Element, or Name of Process). The team should describe a Failure effect from the customer's point of view. Failures that could impact safety or cause non-compliance should be identified clearly in the PFMEA.

Customers could be:

Internal customer (next operation/subsequent operation/operation tar-gets)
External customer (Next Tier Level/OEM/dealer)
Legislative bodies
Product or Product end-user/operator

Failure Effects are given a Severity rating according to:

"Your plant": the effect of the failure mode assuming the defect is detected in the plant (what action will the plant take, e.g., scrap)
"Ship-to plant": the effect of the failure mode assuming the defect is not detected before shipping to the next plant (what action will the next plant take, e.g., sort)
"End-user": the effect of the process item effect (what will the end-user notice, feel, hear, smell, and so on, like, "seat does not move")

The team should ask the following questions to help determine the potential impact of failure effects:

Does the failure mode physically impact downstream processing or cause potential harm to equipment or operators?

This includes an inability to assemble or join a mating component at any subsequent customer's facility.

If so, identify the manufacturing impact "Your plant" and/or "ship-to plant" in the PFMEA. If not, then go to question 2.

Examples could include:

Unable to assemble at operation x
Unable to attach at the customer facility
Unable to connect at the customer facility
Cannot bore at operation X
Causes excessive tool wear at operation X
Damages equipment at operation X
Endangers operator at the customer facility

When parts cannot be assembled there is no impact to the End User and question 2 does not apply.

What is the potential impact on the End-user?

Independent of any controls planned or implemented, including error or mistake-proofing, consider what happens to the process item that leads to what the End User would notice or experience. This information may be available within the DFMEA. If an effect is carried from the DFMEA, the description of the product effects in the PFMEA should be consistent with those in the corresponding DFMEA.

In some cases, the analysis team may not know the end-user effect (e.g., catalog parts, off-the-shelf products, Tier 3 components). When this information is unknown, the effects should be defined in the part function and/or process specification.

Examples could include:

Noise
High effort
Unpleasant odor
Intermittent operation
Water leak
Rough idle
Unable to adjust
Difficult to control
Poor appearance
Regulatory System Function reduced or failed
End-user lack of vehicle control
Safety effect on end-user

What would happen if a failure effect was detected prior to reaching the end-user?

The failure effect at the current or receiving locations also needs to be considered.

Identify the manufacturing impact "Your Plant and/or "ship-to plant" in the PMEA.

Examples could include:

Line shutdown
Stop shipment
Yard hold
100% of products scrapped
Decreased line speed
Increased staffing to maintain the required line rate
Rework and repair

Severity

Severity is a rating number associated with the most severe effect for a given failure mode for evaluating the process step. It is a relative rating within the scope of the individual FMEA and is determined without regard for Occurrence or Detection.

For process-specific effects, the Severity rating should be determined using the criteria in evaluation Table P1. The table may be augmented to include corporate or product line-specific examples.

The customer and the organization should mutually agree to the evaluations of the Failure Effects.

If the customer impacted by a Failure Mode is the subsequent manufacturing or assembly plant or the product user, assessing the severity may lie outside the immediate process engineer’s/team’s field of experience or knowledge. In these cases, the Design FMEA, design engineer, and/or subsequent manufacturing or assembly plant process engineer should be consulted to comprehend the propagation of effects.

Process General Evaluation Criteria Severity (S)
Potential Failure Effects rated according to the criteria below.
S	Effect	Impact to Your Plant	Impact to Ship-to Plant (when known)	Impact to End User (when known)
10	High	Failure may result in an acute health and/or safety risk for the manufacturing or assembly worker	Failure may result in an acute health and/or safety risk for the manufacturing or assembly worker	Affects safe operation of the vehicle and/or other vehicles, the health of driver or passenger(s) or road users or pedestrians.
9	High	Failure may result in in-plant regulatory noncompliance	Failure may result in in-plant regulatory noncompliance	Noncompliance with regulations.
8	Moderatly high	100% of production run affected may have to be scrapped. Failure may result in in-plant regulatory noncompliance or may have a chronic health and/or safety risk for the manufacturing or assembly worker	Line shutdown greater than full production shift; stop shipment possible; field repair or replacement required (Assembly to End User) other than for regulatory noncompliance. Failure may result in in-plant regulatory noncompliance or may have a chronic health and/or safety risk for the manufacturing or assembly worker.	Loss of primary vehicle function necessary for normal driving during expected service life.
7	Moderatly high	Product may have to be sorted and a portion (less than 100%) scrapped; deviation from primary process; decreased line speed or added manpower	Line shutdown from 1 hour up to full production shift; stop shipment possible; field repair or replacement required (Assembly to End User) other than for regulatory noncompliance.	Degradation of primary vehicle function necessary for normal driving during expected service life.
6	Moderatly low	100% of production run may have to be reworked off line and accepted	Line shutdown up to one hour	Loss of secondary vehicle function.
5		A portion of the production run may have to be reworked off line and accepted	Less than 100% of product affected; strong possibility for additional defective product; sort required; no line shutdown	Degradation of secondary vehicle function.
4		100% of production run may have to be reworked in station before it is processed	Defective product triggers significant reaction plan; additional defective products not likely; sort not required	Very objectionable appearance, sound, vibration, harshness, or haptics.
3	Low	A portion of the production run may have to be reworked in-station before it is processed	Defective product triggers minor reaction plan; additional defective products not likely, sort not required	Moderately objectionable appearance, sound, vibration, harshness, or haptics.
2	Low	Slight inconvenience to process, operation, or operator	Defective product triggers no reaction plan; additional defective products not likely; sort not required; requires feedback to supplier	Slightly objectionable appearance, sound, vibration, harshness, or haptics.
1	Very low	No discernible effect	No discernible effect or no effect	No discernible effect

Failure Mode of Process Step

A (Process) Failure Mode is defined as how the process could cause the product not to deliver or provide the intended function.

The team should assume that the product's basic design is correct; however, if design issues result in process concerns, the design issues should communicate those issues to the design team for resolution.

Assume that the failure mode could occur but may not necessarily occur. Failure modes should be described in technical terms, not as a symptom noticeable by the customer.

Verification of completeness of the failure modes can be made through a review of past things-gone-wrong, reject or scrap reports, and group brainstorming. Sources for this should also include a comparison of similar processes and a review of the customer (end-user and subsequent operation) claims relating to similar components.

There are several categories of potential failure modes, including:

Loss of process function/operation not performed
Partial function - Incomplete operation
Degradation of process function
Overachieving process function - Too much too high.
Intermittent process function - operation not consistent
Unstable operation
Unintended process function - wrong operation
Wrong part installed
Delayed process function - operation too late

Typical failure modes could be, but are not limited to:

Hole too shallow, too deep, missing or off location.
Dirty surface
The surface finish is too smooth
Misaligned connector pins
The connector is not fully seated
Pass a lousy part, reject a good part, bypass inspection operation
Label missing
Barcode not readable
ECU flashed with the wrong software.

Failure Cause of the Work Element

A failure cause is an indication of why a failure mode could occur. The consequence of a cause is the failure mode. To the extent possible, identify every potential manufacturing or assembly cause for each failure mode. The cause should be listed as wholly and concisely as possible so that efforts (controls and actions) can be aimed at appropriate causes.

Typical failure causes may include the classic Ishikawa’s 4M, but are not limited to:

Man: set-up worker, machine operator associate, material associate, maintenance technician, etc.
Machine/Equipment: robot, hopper reservoir tank, injection molding machine, spiral conveyor, inspection devices, fixtures, etc.
Material (Indirect): machining oil, installation grease, washer concentration, (aid for operation), etc.
Environment (Milieu): ambient conditions such as heat, dust, contamination, lighting, noise, etc.

In preparing the FMEA, assume that the incoming part(s)/material(s) are correct. Exceptions can be made by the FMEA team where historical data indicate deficiencies in incoming part quality.

One method to help reveal/uncover failure causes is having a facilitator who leads the team through "Thought-Provoking Stimulation Questions." These questions can be broad category questions, enough to stimulate the process experts thought process, while keeping the number of questions manageable. Questions can be process-specific and broken down into the 4M categories. A review of previous PFMEAs can form an initial list of questions.

Example - Assembly Process:

Man

From parts available within the process, can the wrong part be applied?
Can no part be applied?
Can the parts be loaded incorrectly?
Can parts be damaged - From pickup to application?
Can the wrong material be used?

Machine

Can automated process be interrupted?
Can inputted data be entered incorrectly?
Can machine be run in manual mode, bypassing automated controls?
Is there a schedule to confirm prevention and detection controls?

Material (indirect)

Can too much / too little / no material be used?
Can material be applied to a wrong location?

EnvironMent (Milieu)

Is lighting adequate for task?
Can parts used within the process, be considered foreign material?

The description of the failure cause needs to be clear. Terms such as "defective, broken," "operator failure," non-fulfillment or “not OK" and so on are insufficient to comprehensively assign the failure cause and mode and to determine actions.

Risk analysis (step 5)

Current prevention control (PC) of FC

Current prevention should be implemented and effective. Confirmation of control status can be done during an in-station review (e.g., Line Side Review, Line walks, and Regular audits). If the control is not adequate, additional action may be needed.
The Occurrence and Detection ratings should be reviewed when using data from previous processes due to the possibility of different conditions for the new process.

Evaluations

Each Failure Mode, Cause and Effect relationship (failure chain or net) independent risk is assessed. There are three rating criteria for the evaluation of risk:

Severity (S): stands for the Severity of the Failure Effect
Occurrence (O): stands for the Occurrence of the Failure Cause
Detection (D): stands for the Detection of the occurred Failure Cause or Failure Mode

Evaluation numbers from 1 to 10 are used for S, O, and D, respectively, in which 10 stands for the highest risk contribution.

It is not suitable to compare the ratings of one team’s FMEA with the ratings of another team’s FMEA, even if the product/process appears to be identical.

Occurrence

The Occurrence rating (0) describes the occurrence of Failure Cause in the process, taking into account the associated current prevention controls.

The occurrence rating number is a relative rating within the scope of the FMEA and may not reflect the actual occurrence.

The Occurrence rating describes the potential of the failure cause to occur without regard to the detection controls.

Expertise or other experiences with similar processes, for example, can be considered in the assessment of the rating numbers.

In determining this rating, the team should consider questions such as the following:

What is the equipment history with similar processes and process steps?
What is the field experience with similar processes?
Is the process a carryover or similar to a previous process?
How significant are changes from a current production process?
Is the process completely new?
What are the environmental changes?
Are best practices already implemented?
Do standard instructions exist? (e.g., work instructions, set-up and calibration procedures, preventive maintenance, error proofing verification procedures, and process monitoring verification checklists)
Are technical error-proofing solutions implemented? (e.g., product or process design, fixture and tool design, established process sequence, production control tracking/traceability, machine capability, and SPC charting)

Occurrence Potential (O) for the Process
Potential Failure Causes rated according to the criteria below. Consider Prevention Controls when determining the best Occurrence estimate. Occurrence is a predictive qualitative rating made at the time of evaluation and may not reflect the actual occurrence. The occurrence rating number is a relative rating within the scope of the FMEA (process being evaluated). For Prevention Controls with multiple Occurrence Ratings, use the rating that best reflects the robustness of the control
O	Prediction of Failure Cause Occuring	Type of Control	Prevention Controls
10	Extremely high	None	No prevention controls
9	Very high	Behavioral	Prevention controls will have little effect in preventing failure cause.
8	Very high	Behavioral
7	High	Behavioral or Technical	Prevention controls somewhat effective in preventing failure cause
6	High
5	Moderate		Prevention controls are effective in preventing failure cause.
4	Moderate
3	Low	Best Practices: Behavioral or Technical	Prevention controls are highly effective in preventing failure cause
2	Very low	Best Practices: Behavioral or Technical
1	Extremely low	Technical	Prevention controls are extremely effective in preventing failure cause from occurring due to design (e.g. part geometry) or process (e.g.fixture or tooling design). Intent of prevention controls - Failure Mode cannot be physically produced due to the Failure Cause.

Current detection control (DC) of FC

Current detection should be implemented and effective. Confirmation of control status can be done during an in-station review (e.g., Line Side Review, Line walks, and Regular audits). If the control is not adequate, additional action may be needed.
The Occurrence and Detection ratings should be reviewed when using data from previous processes due to the possibility of different conditions for the new process.

Detectability

Detection is the rating associated with a prediction of the most effective process control from the listed detection-type process controls. Detection is a relative rating within the scope of the individual FMEA and is determined without regard for Severity or Occurrence. Detection should be estimated using the criteria in Table P3. The company may extend this table with examples of standard detection methods used in their environment.

The intent of the term "control discrepant product" used in Table P3 Ranks 3 and 4 is to have controls/systems/procedures in place that control the discrepant product so that the probability of the product escaping the facility is very low. The controls start from when the product is identified as discrepant to the point of final disposition. These controls usually exceed controls that are used for discrepant products with higher Detection Ranks.

After implementing any unproven control, the effectiveness can be verified and re-evaluated.
While estimating Detection, the team should consider questions such as the following:

Which test is most effective in detecting the Failure Cause or the Failure Mode?
What is the usage Profile / Duty Cycle required to detect the failure?
What sample size is necessary to detect the failure?
Is the test procedure proven for detecting this Cause/Failure Mode?

Detection Potential (D) for the Validation of the Process Design
Detection Controls rated according to Detection Method Maturity and Opportunity for Detection
D	Ability to Detect	Detection Method Maturity	Opportunity for Detection
10	Very low	No testing or inspection method has been established or is known	The failure mode will not or cannot be detected.
9	Very low	It is unlikely that the testing or inspection method will detect the failure mode	The failure mode is not easily detected through random or sporadic audits.
8	Low	Test or inspection method has not been proven to be effective and reliable (e.g. plant has little or no experience with method, gauge R&R results marginal on comparable process or this application, etc.).	Human inspection (visual, tactile, audible), or use of manual gauging (attribute or variable) that should detect the failure mode or failure cause.
7	Low		Machine-based detection (automated or semi-automated with notification by light, buzzer, etc.), or use of inspection equipment such as a coordinate measuring machine that should detect failure mode or failure cause.
6	Moderate	Test or inspection method has been proven to be effective and reliable (e.g. plant has experience with method; gauge R&R results are acceptable on comparable process or this application, etc.).	Human inspection (visual, tactile, audible), or use of manual gauging (attribute or variable) that will detect the failure mode or failure cause (including product sample checks).
5	Moderate		Machine-based detection (semi-automated with notification by light, buzzer, etc.), or use of inspection equipment such as a coordinate measuring machine that will detect failure mode or failure cause (including product sample checks).
4	High	System has been proven to be effective and reliable (e.g. plant has experience with method on identical process or this application), gauge R&R results are acceptable, etc.	Machine-based automated detection method that will detect the failure mode downstream, prevent further processing or system will identify the product as discrepant and allow it to automatically move forward in the process until the designated reject unload area. Discrepant product will be controlled by a robust system that will prevent outflow of the product from the facility
3			Machine-based automated detection method that will detect the failure mode in station, prevent further processing or system will identify the product as discrepant and allow it to automatically move forward in the process until the designated reject unload area. Discrepant product will be controlled by a robust system that will prevent outflow of the product from the facility
2		Detection method has been proven to be effective and reliable (e.g plant has experience with method, error-proofing verifications, etc.).	Machine-based detection method that will detect the cause and prevent the failure mode (discrepant part) from being produced
1	Very high	Failure mode cannot be physically produced as-designed or processed, or detection methods proven to always detect the failure mode or failure cause.

Action priority

The Action Priority (AP) method accounts for all 1000 possible combinations of S, O, and D. It emphasizes, in order, severity, occurrence, and detection. It follows the failure-prevention intent of FMEA. The AP table offers a suggested high-medium-low priority for action. Companies can use a single system to evaluate action priorities instead of multiple systems required from numerous customers.

Risk Priority Numbers are the product of S x Ox D and range from 1 to 1000. The RPN distribution can provide some information about the range of ratings. Still, RPN alone is not an adequate method to determine the need for more actions and is not a recommended practice to determine the need for action. RPN gives equal weight to S, O, and D. Hence, it yields similar risk numbers for very different combinations of S, O, and D leaving the team uncertain about prioritizing. When using RPN, it is recommended to use an additional method to prioritize RPN results, such as SxO.
Risk matrices can represent combinations of S and O, S and D, and O and D. These matrices provide a visual representation of the analysis results. They can be used as an input to prioritize actions based on company-established criteria.

Since the AP Table works with the Severity, Occurrence, and Detection tables, the AP table also needs a review if the organization chooses to modify the S, O, and D -tables.

Priority High (H):
Highest priority for review and action.
The team needs to either identify appropriate action to
improve prevention or detection controls or justify and
document why current controls are adequate
Priority Medium (M):
Medium priority for review and action. The team should identify appropriate actions to improve prevention or detection controls, or, at the discretion of the company, justify and document why controls are adequate.
Priority Low (L):
Low priority for review and action, The team could identify actions to improve prevention or detection controls.