FAI verifies if a single first-run part meets drawing specifications. PPAP approves the entire manufacturing process for stable serial production. Cpk measures how consistently that specific process stays within critical tolerance limits.
Understanding this difference is the boundary between a successful product launch and an expensive manufacturing disaster. A supplier may deliver a flawless prototype, but one good part cannot guarantee that their tooling, operators, and quality controls will produce the same result 10,000 times without failing.
To ensure quality from the first sample to volume production, buyers use three distinct validation tools. Here is how they compare:
| Quality Tool | Main Purpose | Typical Use | Main Limitation |
|---|---|---|---|
| FAI | Verify part conformity | First production-intent build | Does not prove long-term capability |
| PPAP | Support production approval | Before serial production | Does not replace ongoing control |
| Cpk/Ppk | Measure process capability | Pilot run and production | Requires suitable and stable data |
| SPC | Monitor process stability | Serial production | Does not provide initial approval |
Note: This article provides general manufacturing guidance. Contract terms, customer-specific requirements, and industry standards always define the final approval scope.
FAI Checks the Part; PPAP Approves the Process
Securing a compliant physical part is the first validation step. Proving the reliability of the manufacturing system that built it is the second.
First Article Conformity
FAI shows whether a representative production-intent part matches the approved drawing and specification.
During an FAI, quality engineers systematically verify:
- Dimensions and tolerances
- GD&T and datums
- Material requirements
- Technical notes
- Thread and assembly features
- Surface finish
- Appearance and functional requirements
An FAI does not necessarily inspect the absolute first part produced by the equipment. The sample typically comes from the initial production-intent run and is selected according to customer-specified sampling methods.
A reviewable FAI report requires actual measured values. Recording “OK” or “Pass” fails to show how much margin remains before a feature drifts out of tolerance. For our steel bracket, recording a bend angle as “89.5°” provides actionable data; recording it as “Pass” does not.
Production Process Approval
PPAP shows whether the supplier has defined and controlled the process needed for repeat production.
Depending on the submission level, a PPAP package may include:
- Design records
- Process flow
- PFMEA (Process Failure Mode and Effects Analysis)
- Control plan
- MSA (Measurement System Analysis) or GR&R
- Dimensional results
- Material and performance tests
- Initial capability studies
- Traceability records
- Part Submission Warrant (PSW)
PPAP is a key output of the Advanced Product Quality Planning (APQP) framework. While APQP manages the broader product quality lifecycle, PPAP serves as the specific evidence required for production authorization. For the bracket, PPAP ensures the supplier has a control plan for powder coating thickness, preventing rust issues in month six.
| Party | Main Responsibility |
|---|---|
| Customer | Defines CTQs, forms, submission level, and acceptance rules |
| Supplier | Produces parts and prepares the required evidence |
| Customer Approval Team | Reviews, approves, temporarily approves, or rejects the submission |
Where Cpk Fits
Cpk provides statistical evidence for selected measurable characteristics. It does not approve the complete part or process by itself.
To understand how these validation tools interact:
- FAI usually verifies all drawing requirements.
- Cpk/Ppk is generally applied only to selected Critical-to-Quality (CTQ) features.
- Capability studies often serve as one supporting document within a PPAP submission.
- Passing an FAI does not prove the manufacturing process is capable.
- An acceptable Cpk on a specific hole diameter does not prove the powder coating adhesion, material grade, or overall drawing requirements are met. Cpk is perfect for checking the PEM fastener hole diameter stability, but it cannot verify the cosmetic finish.
Industry context matters here. Automotive projects usually rely on customer-mandated PPAP frameworks. Aerospace, medical, and general industrial sectors may utilize different validation methods, such as AS9102 for aerospace first articles, paired with separate process validation protocols.
From First Sample to Serial Production
Quality documentation requirements align with specific phases of the manufacturing lifecycle. This is not a rigid sequence, but a general progression of risk reduction.
Production-Intent Parts
It is important to distinguish between a prototype and a production-intent part. Prototypes are primarily used to verify design, fit, and manufacturing feasibility. They often rely on temporary tooling, alternative equipment, or extra manual adjustments.
A production-intent part must be manufactured using:
- Planned material
- Planned equipment
- Formal or near-formal tooling
- Planned operational sequences
- Formal inspection methods
- The current drawing revision
For our mounting bracket, a prototype might be folded on a manual press brake. The production-intent part must be bent on the automated CNC press brake using the final production dies. FAI samples can be pulled from this production-intent run or another batch as directed by the customer.
Pilot Run and PPAP Submission
The pilot run exposes hidden manufacturing issues under conditions that closely mimic normal volume production.
A pilot run specifically tests:
- Equipment and tooling stability
- Tool and die wear rates
- Operator repeatability
- Planned production cycle times
- Inspection methods and fixture accuracy
- Packaging and handling procedures
- Batch traceability systems
- CTQ data collection
FAI dimensional results and initial Cpk/Ppk data are typically extracted from this pilot batch. The exact sample size and sampling method depend on the customer, the process stability, and the part’s risk profile. There is no universally fixed standard requiring exactly 50, 100, or 200 parts; the volume dictates the appropriate scale for the run.
Buyer’s Warning: If the pilot run relies on 100% manual sorting to remove defective brackets just to pass the capability requirements, the process is not ready for serial production, and the PPAP should be rejected.
Serial Production Control
Once PPAP is approved, the process enters controlled serial production.
Continuous control during this phase usually includes:
- SPC and control charts
- First-off and patrol inspections
- Batch sampling
- Tooling and fixture maintenance
- Equipment parameter monitoring
- Material traceability
- Non-conformance reaction plans
- Periodic capability reviews
PPAP establishes the approval baseline, but it does not replace the need for ongoing process monitoring. For the steel bracket, this means continuous monitoring of the laser cutting gas pressure, routine validation of the bending dies, and regular pull-out tests on the PEM fasteners long after the initial PPAP warrant is signed.
A Reviewable FAI and PPAP Package
Throwing a dozen unrelated PDFs into a ZIP file does not constitute a valid submission. A reviewable approval package presents interconnected evidence. The dimensional report must match the control plan, which must align with the manufacturing flow.
Drawing and Dimensional Results
The dimensional report is the foundation of the FAI. It requires systematic documentation of every feature on the print.
A complete dimensional package includes:
- Current drawing revision
- Ballooned drawing
- Nominal values
- Tolerance ranges
- Actual measured values
- Pass/Fail status
- Inspection equipment used
- Datum establishment methods
- Fixturing and measurement conditions
- Part and report identification numbers
For our steel bracket, if the drawing calls out 45 dimensions, the FAI must trace 45 actual measured values back to the corresponding ballooned drawing.
| Weak FAI | Reviewable FAI |
|---|---|
| Shows only “OK” | Shows actual measured values |
| Drawing revision is unclear | Matches the approved drawing revision |
| Measurement method is missing | Lists the gauge, CMM, or fixture |
| CTQs are not identified | Separates CTQs from routine dimensions |
| Results cannot be traced | Links every result to a balloon number |
Process and Measurement Records
To approve the process, the buyer must review the documentation that governs how the part is made and verified on the shop floor.
Process records typically include:
- Process flow diagram
- PFMEA (Process Failure Mode and Effects Analysis)
- Control plan
- Work instructions
- MSA (Measurement System Analysis) or GR&R
- Cpk or Ppk results
- Inspection frequencies
- Reaction plans
- Approved process settings
Measurement equipment must match the tolerance requirements of the specific feature. Calipers work well for the bracket’s overall flat blank length. Pin gauges are appropriate for checking the pre-plated PEM insertion holes. A CMM is required to verify the true position of the mounting holes relative to the primary datum.
Buyer’s Warning: Different measurement methods yield different results. If the supplier measures the bracket resting freely on a surface plate, and your incoming inspection measures it clamped in a fixture, the dimensional results will conflict. Datums, fixturing, and measurement methods must be aligned prior to the FAI submission.
Material and Traceability Records
Material failures often remain hidden until the part is deployed in the field. Depending on the project scope, traceability records may include:
- Material Test Reports (MTR) or material certificates
- Certificate of Conformance (CoC)
- Heat treatment certifications
- Surface finish certifications
- Raw material batch numbers
- Sub-tier supplier records
- Production and inspection batch numbers
- ECO (Engineering Change Order) records
- Packaging and shipping logs
Traceability connects the final assembly back to its raw components. If a bracket suffers from premature rust or a PEM nut pulls out during final assembly, the lot number on the box must allow you to trace back to the exact sub-tier plating batch, the specific powder coating run, and the original steel master coil.
Without strict lot traceability, a single field failure might force you to quarantine and scrap your entire inventory. With it, you only contain the specific suspect batch, saving thousands of dollars in isolation costs.
Cpk in Practice: Read the Process, Not Just the Score
Capability indices are not just compliance scores; they predict the likelihood of future defects. You do not need to memorize the statistical formulas to understand what the data tells you about the manufacturing process.
Cp, Cpk, Pp, and Ppk
Think of process capability like parking a car in a garage.
- Cp compares the width of your car to the width of the garage. It tells you if the car can physically fit.
- Cpk checks your centering. Even if the car is narrow enough, parking too far to the left means you will scrape the wall. Cpk accounts for both the process spread and its alignment with the target.
- Pp and Ppk apply the same logic but use the overall, long-term variation of the process, reflecting how you park every day over a month rather than on a single good afternoon.
| Index | Main Focus |
|---|---|
| Cp | Short-term process spread |
| Cpk | Short-term spread and centering |
| Pp | Overall process spread |
| Ppk | Overall spread and centering |
Whether you require Cpk, Ppk, or both depends on your specific quality standards and the nature of the data collected during the pilot run.
Stable Data and Measurement Confidence
A Cpk score is only as reliable as the data behind it. Before calculating capability, you must confirm:
- The process is in a state of statistical control.
- Data is recorded in exact production order.
- Parts were produced under normal operating conditions.
- Different machines, shifts, or mold cavities were not mixed into a single data set.
- Out-of-spec parts were not deleted from the record.
- Gages are calibrated and passed MSA/GR&R.
- The data distribution fits the statistical method used (e.g., normal distribution).
Beginners often confuse specification limits with control limits:
- USL and LSL (Specification Limits): Come directly from the engineering drawing. They define what the customer accepts.
- UCL and LCL (Control Limits): Come from the statistical process data. They define what the manufacturing process naturally does.
You cannot use specification limits to calculate control charts, and control limits cannot override the drawing tolerances.
Furthermore, traditional Cpk only works for continuous, measurable variables (like the inner diameter of a hole). You cannot calculate Cpk on attribute data, such as the cosmetic appearance of the bracket’s powder coat or a go/no-go thread check. Those require defect rate tracking or attribute capability analysis.
Capability Targets and Reaction Plans
| Cpk Result | General Interpretation |
|---|---|
| Below 1.00 | Process spread or centering does not fit the specification. High risk of defects. |
| 1.00–1.33 | Limited margin. Greater sensitivity to normal process variation. |
| 1.33 or above | A common customer target for standard critical characteristics. |
| 1.67 or above | Often required for new, automotive, or higher-risk characteristics. |
| 2.00 or above | Strong statistical margin (Six Sigma quality level), though not a zero-defect guarantee. |
These values serve as general references. Actual approval criteria are dictated by customer requirements, industry standards, and the specific risk of the feature.
If a supplier accepts a process with a Cpk below 1.0, they cannot naturally meet the tolerance. Ultimately, you are paying for their hidden scrap rates and 100% manual sorting costs baked into the piece price.
When a CTQ yields an insufficient Cpk (e.g., the bracket’s bend angle scores a 0.85), the supplier must implement reaction plans to protect the customer. Common reactions include:
- Instituting 100% sorting or inspection for that specific feature.
- Increasing the sampling frequency.
- Adjusting the manufacturing process (e.g., modifying the press brake parameters).
- Upgrading tooling or fixturing.
- Isolating data to find the root cause (e.g., separating data from Machine A vs. Machine B).
- Submitting a deviation request for a temporary tolerance concession.
- Conducting a new capability study after corrective actions are verified.
Buyer’s Warning: Note that 100% manual inspection is a temporary band-aid, not a cure. Human sorting is notoriously prone to error and fatigue. Long-term PPAP approval should require actual process improvements, such as tooling upgrades or parameter optimization, rather than relying on operators to act as the final quality filter.
Set the Approval Scope Before Quotation
Quality documents require labor, machine time, and specialized equipment. Demanding a full PPAP for a 50-piece prototype run is a waste of capital, while accepting a basic dimensional report on a 50,000-piece critical component is a massive supply chain risk. You must align the documentation scope with the project risk before the supplier quotes the job.
Risk-Based Document Scope
There is no one-size-fits-all approach to quality documentation. The table below outlines possible evidence scopes based on typical manufacturing scenarios.
| Project Situation | Possible Evidence Scope |
|---|---|
| Prototype and design learning | Basic dimensional report and material reference |
| Low-volume build-to-print part | FAI and material certificate |
| Repeat production with CTQs | FAI, control plan, and capability evidence |
| Customer-mandated serial program | Required PPAP submission (e.g., Level 3) |
| Regulated or safety-related part | Contractually required validation and traceability |
| Established repeat order | Batch report, CoC, and ongoing control records |
We use the term “possible evidence scope” because industry standards and customer requirements dictate the final package.
- Automotive projects generally default to the AIAG PPAP framework.
- Aerospace contracts frequently mandate AS9102 FAI reports.
- Medical device manufacturing requires strict process validation (IQ/OQ/PQ), traceability, and regulatory compliance files.
Ultimately, the buyer’s purchase order and Supplier Quality Agreement (SQA) define what is required. A custom steel bracket used in a server rack requires different evidence than the exact same bracket used in a hospital ventilator.
RFQ and Acceptance Criteria
If you do not specify quality deliverables in the Request for Quote (RFQ), the supplier will not budget the labor to execute them. Requesting a 30-piece capability study after the price is locked will inevitably cause friction over added costs, and more importantly, delay your product launch by weeks.
Your RFQ should explicitly define:
- The requirement for an FAI, partial PPAP, or full PPAP submission.
- The required PPAP submission level (Level 1 through 5).
- Whether the supplier must use your specific customer templates or their own standard forms.
- Identified CTQs and Special Characteristics.
- Target Cpk or Ppk minimums (e.g., > 1.33).
- The pilot run sample size and specific selection stages.
- Required measurement equipment and fixturing states (e.g., measured clamped vs. free state).
- Material, heat treatment, and special process certificates required.
- Submission deadlines.
- Who holds the audit and approval responsibility.
- The rules for deviation approvals.
These requirements directly impact the supplier’s cost structure. They dictate CMM programming time, the number of brackets scrapped for destructive PEM pull-out testing, inspection labor hours, and the overall production lead time. Upfront clarity prevents budget surprises.
Changes That Require Reapproval
An approved FAI or PPAP is a snapshot of a specific process at a specific time. If that process changes, the approval baseline is void.
The following events typically trigger a mandatory resubmission of a partial FAI, full FAI, or PPAP:
- Engineering drawing or tolerance changes.
- Material grade or raw material source changes.
- Tooling modifications, repairs, or replacements (e.g., grinding the press brake die).
- Equipment replacement or relocation within the plant.
- Facility relocation.
- Changes to sub-tier suppliers (e.g., switching to a new powder coating vendor).
- Adjustments to critical process parameters.
- Resuming production after an extended shutdown period.
After reviewing a resubmission, the customer approval team will issue a status:
- Approved: The part is cleared for normal serial production.
- Interim Approval: Permits shipment of material for production requirements on a limited time or piece quantity basis, usually while minor documentation issues are resolved. (Buyer’s note: Never grant an interim approval without a hard expiration date or strict quantity limit, or the root cause will never be fixed.)
- Rejected: The submission does not meet requirements; production shipment is not authorized.
- Resubmission Required: Specific elements must be corrected and reviewed again.
Buyer’s Warning: Suppliers must not unilaterally decide that a change “doesn’t affect the part” and skip notifying the customer. The specific triggers for reapproval must be legally bound in your procurement requirements or quality agreements.
Conclusion
- FAI confirms that a production-intent part meets the drawing.
- PPAP provides the evidence needed to approve repeat production.
- Cpk measures the capability of a stable process on selected measurable characteristics.
Send your drawing, CTQ list, expected volume, and approval requirements before requesting a quote. This allows the supplier to plan the correct inspection, capability study, and submission scope before production begins. Looking for a manufacturing partner who doesn’t guess on quality? Contact our engineering team today to build a validation plan that fits your exact risk profile and budget.
Hey, I'm Kevin Lee
For the past 10 years, I’ve been immersed in various forms of sheet metal fabrication, sharing cool insights here from my experiences across diverse workshops.
Get in touch
Kevin Lee
I have over ten years of professional experience in sheet metal fabrication, specializing in laser cutting, bending, welding, and surface treatment techniques. As the Technical Director at Shengen, I am committed to solving complex manufacturing challenges and driving innovation and quality in each project.
