Dfmea Design

Design FMEA (DFMEA) — AIAG-VDA 2019

Goal

Identify and mitigate design risks by analysing functions, interfaces, and failure modes — so that design weaknesses are resolved before release to manufacturing, and the results feed directly into the DVP, PFMEA, and Control Plan.

Required Execution Checklist

• DFMEA scope defined: boundary diagram complete, inside/outside boundary identified
• Product hierarchy mapped: system → subsystem → component → interface (Step 2)
• All interfaces identified — internal (component-to-component), external (environment), and user/assembly
• All functions defined with Verb + Noun + Measurable Standard (Step 3)
• Each function has a measurable verification method confirmed in the DVP
• Failure chain complete: Effect → Mode → Cause for each function (Step 4)
• All Failure Causes validated using engineering analysis (FEA, calculation) or test evidence — not assumptions
• S/O/D ratings justified using analysis, test data, or design history
• All H-AP items have a defined action, named owner, and target date
• DFMEA → DVP, PFMEA, and drawing linkage verified before release to manufacturing

When to use

• New product development (integrate into APQP Phase 2 — Product Design and Development)
• Design change or engineering change request (ECR)
• Field failure investigation revealing a design root cause
• Periodic design review
• Before handoff to manufacturing (DFMEA drives the PFMEA)

Key difference from PFMEA: DFMEA analyses the design intent and design robustness. PFMEA analyses the manufacturing process. DFMEA comes first — its failure effects and severity ratings inform the PFMEA.

Prerequisites

• Product requirements / specification (engineering drawing, customer spec)
• System block diagram or product breakdown structure
• Interface matrix (if system-level analysis)
• Team: design engineer, systems engineer, quality engineer, reliability (if available)

The 7-Step AIAG-VDA 2019 Approach for DFMEA

Step 1 — Planning and Preparation

Define scope:

• Analysis object: component, subsystem, or system
• Customer: who is the next-level assembly? Who is the end user?
• Boundary diagram: what is inside and outside the DFMEA scope
• Interface matrix: what interacts with this component (mechanical, electrical, thermal, chemical)?

Step 2 — Structure Analysis (Design Hierarchy)

Map the product hierarchy:

System (e.g., Steering Column)
└── Subsystem (e.g., Tilt Mechanism)
    └── Component (e.g., Pivot Pin)
        └── Interface (e.g., Pin-to-Bracket contact)

Interface matrix: For each component, identify:

• Interfaces to other components (internal)
• Interfaces to the environment (external): heat, vibration, corrosion, electromagnetic
• Interfaces to the user or assembly process

Interfaces are where most design failures occur. Each interface must be analysed as a potential failure location in its own right — not just the components that share it. In practice, at least 50% of DFMEA effort should focus on interfaces and interactions; component-only analysis misses the most common field failure modes.

Step 3 — Function Analysis

For each element in the structure, define its design function:

Format: Verb + Noun + Measurable Standard

Examples:

• Component function: "Transmit torque of 50 Nm ± 5 Nm without permanent deformation"
• Interface function: "Maintain sealing at pressure 2.5 bar across -40°C to +120°C"
• System function: "Provide steering angle feedback with latency < 50ms"

Identify Special Characteristics from the drawing — these get S = 9 or 10 in Step 5.

Testability rule: Every function must be measurable and verifiable — a test or analysis method must exist in the DVP. A function with no verification method cannot receive a credible D rating in Step 5.

Step 4 — Failure Analysis

The failure chain for DFMEA: Failure Effect → Failure Mode → Failure Cause

Failure Effect (FE):

• End-user effect: safety hazard, loss of primary function, reduced performance
• Vehicle/system effect: damage to adjacent components, secondary failures
• Manufacturing effect (if component is not to spec): inability to assemble, rework

Failure Mode (FM): How does this component fail to perform its design function?

• Fracture, wear, corrosion, deformation, signal loss, leakage, dimensional drift, electrical short/open

Failure Cause (FC): What design parameter or design decision causes the failure mode?

• Insufficient material strength (wrong grade, wrong heat treat)
• Inadequate geometry (stress concentration, insufficient wall thickness)
• Thermal expansion mismatch
• Corrosion protection insufficient for environment
• Tolerance stack-up causing interference
• Interface design not accounting for assembly variation

Failure Cause validation: Failure Causes must be validated using engineering analysis (FEA, structural calculation, thermal simulation) or test evidence. Unverified assumptions — "probably," "likely," "may be" — are not acceptable as final Failure Causes. Use CAE (FEA, simulation), test data, and field return data to support both the failure modes identified and their causes. For post-field-failure DFMEA updates, the FC must match the validated root cause from the 8D or field investigation.

Step 5 — Risk Analysis

Use the same S/O/D ratings and AP table as PFMEA (see action-priority-ap and pfmea-process for the tables).

Key DFMEA-specific notes:

Occurrence (O) in DFMEA considers:

• Design maturity (new design = higher O)
• Similar design history (similar designs with this failure = higher O)
• Prevention controls: design reviews, CAE / FEA, DVP testing

Detection (D) in DFMEA considers:

• Design verification tests (DVP)
• Engineering analysis (FEA, simulation)
• Design reviews with checklists
• Prototype testing

D = 10 means no design verification for this failure mode. This is common for new features — it drives the DVP.

Ratings justification: All S/O/D ratings must be justified using analysis, test data, or design history. A rating without documented justification will not withstand OEM audit. Where design history exists from similar components, reference it explicitly. Where data is absent, mark O and D as preliminary and flag the DVP entry that will confirm them.

Step 6 — Optimization

For H-AP items in DFMEA, actions typically fall into:

1. Design change: change geometry, material, tolerance, or coating
2. Add design verification: add DVP test entry to confirm the design meets the requirement
3. Add prevention: design guideline, DFM/DFA rule, standard reference
4. Improve detection: add inspection requirement to DVP or design review checklist

DFMEA → DVP linkage: Every detection control (D rating improvement) must have a corresponding entry in the Design Verification Plan (DVP). If you improve D because "we will test it," the test must exist in the DVP.

Action tracking: All actions must be tracked to closure with objective evidence — test results, simulation output, or updated analysis. Actions marked "complete" without verification evidence are not acceptable. Open H-AP items past their target date must be escalated to the design review owner or programme manager, with a revised date and documented reason for delay.

Step 7 — Results Documentation

DFMEA outputs that feed other APQP documents:

• Special Characteristics identified in DFMEA → transferred to Control Plan and PFMEA
• Detection controls → DVP entries
• Interface failure modes → PFMEA structure analysis inputs
• Material / geometry requirements confirmed → released drawing

DFMEA → PFMEA handoff

The DFMEA and PFMEA are linked:

Mandatory DFMEA review events (APQP / IATF 16949 §8.3)

A DFMEA that is not updated through the programme lifecycle is a design quality assurance gap — not a living document.

Common mistakes

• Starting DFMEA after design is frozen — DFMEA has no value if no changes can be made
• Only analysing the component, not the interfaces — most design failures are interface failures
• Improving D without adding DVP tests — detection credit without actual testing
• DFMEA and PFMEA teams working in silos — they must share severity ratings and special characteristics

Output Format

At the start of each use, ask the user:

"How would you like to receive the output? A — Structured Markdown (formatted tables and sections, ready to copy) B — Plain tables (simplified structure for Excel or Word) C — Narrative report (flowing text for a formal document or email)

Default: A."

Adapt all output sections to the chosen format. If the platform or session context already defines a format preference, skip this question.

Reference files

• Interface matrix construction guide

Changelog

Source: hashgraph-online/awesome-codex-plugins → plugins/RBraga01/Quality-Engineering-Skills/skills/risk-analysis/dfmea-design/SKILL.md