MFG-CNC14-07 · Lesson 11 · 18 clock hours · lab
Statistical Process Control and First-Article Inspection
Objective. Students complete a first-article inspection (FAI) report for a machined part and construct a basic X-bar and range (X̄-R) control chart from a provided data set, identifying whether the process is in statistical control.
⚠ Pending professional review
This lesson, its worked examples and its practice questions has not yet been verified by a licensed engineer or certified journey-level tradesperson or by a certification or approval body. It is a study draft and may contain errors. Do not use it as instructions for work on a job site or on live equipment. Always follow your instructor, your supervising professional, and current official references and regulations.
VoTechWorks program approval and accreditation are still in progress. Credentials are issued only by the certifying bodies named in each program, after their own exams.
Why SPC and First-Article Inspection Matter in CNC Machining
CNC machining can hold tolerances in the thousandths of an inch, but tight tolerances only stay tight when the process is monitored and controlled. Two tools that professionals rely on every day are the first-article inspection (FAI) and statistical process control (SPC). FAI confirms that the very first part produced to a new setup meets every requirement on the print. SPC uses math to watch the process over time and warn you before parts go out of tolerance.
---
First-Article Inspection (FAI)
A first-article inspection is a complete, documented measurement of every dimension and feature called out on the engineering drawing for the first part (or first batch) from a new or revised setup. Standards that govern FAI forms include AS9102 (aerospace) and the automotive PPAP (Production Part Approval Process). Both require you to record:
- The nominal dimension and tolerance from the print
- The actual measured value
- Whether the dimension is in or out of tolerance
- The measurement tool used
- An overall part disposition: Accept or Reject
How to Conduct an FAI
1. Obtain the current revision of the part drawing and the blank FAI form.
2. List every ballooned dimension in order.
3. Measure each feature with the correct instrument (micrometer, CMM, height gage, bore gage, etc.) and record the reading to the precision the tolerance demands.
4. Compare each actual value to the tolerance band; flag any that fall outside.
5. Sign and date the report. Any out-of-tolerance dimension makes the overall disposition Reject — the setup must be corrected and a new FAI run.
Safety note: Deburr and clean the part before handling. Use cut-resistant gloves when handling freshly machined edges. Keep the measuring area clean and dry — coolant on a surface plate causes slipping and measurement error.
---
Statistical Process Control (SPC)
SPC uses control charts to distinguish normal process variation (common cause) from signals that the process has shifted (special cause). The most common chart pair for continuous measurement data is the X̄-R chart (X-bar and Range chart).
Collecting Subgroup Data
A subgroup is a small sample of parts taken together at one point in time — typically three to five consecutive pieces. For each subgroup you calculate:
- X̄ (subgroup mean): add all values in the subgroup, divide by subgroup size n
- R (range): largest value minus smallest value in the subgroup
Control Limit Formulas
Using the standard constants for subgroup size n = 5:
| Constant | Value (n = 5) |
|----------|---------------|
| A₂ | 0.577 |
| D₃ | 0 |
| D₄ | 2.114 |
From 25 subgroups, compute the grand mean X̄̄ and the average range R̄:
- X̄̄ = sum of all subgroup X̄ values ÷ 25
- R̄ = sum of all subgroup ranges ÷ 25
Control limits for the X̄ chart:
- UCL_X̄ = X̄̄ + (A₂ × R̄)
- LCL_X̄ = X̄̄ − (A₂ × R̄)
Control limits for the R chart:
- UCL_R = D₄ × R̄
- LCL_R = D₃ × R̄ = 0 (no lower control limit when n ≤ 6)
Plotting the Chart
Draw two separate graphs on the same sheet: X̄ on top, R on the bottom. Plot the centerline (X̄̄ and R̄) as solid lines and the upper and lower control limits as dashed lines. Plot each subgroup's X̄ and R as a dot and connect them in time order.
Western Electric (WE) Out-of-Control Rules
A process is signaling a special cause if any of these patterns appear:
1. One point beyond either control limit.
2. Two of three consecutive points in the outer one-third of the chart (beyond 2σ on the same side).
3. Four of five consecutive points in the outer two-thirds (beyond 1σ on the same side).
4. Eight or more consecutive points on the same side of the centerline (a run).
Responding to an Out-of-Control Signal
If the X̄ chart shows the process mean drifting upward toward the UCL, the team investigates:
- Tool wear (most common in CNC — the cutter gets dull and the dimension grows)
- Thermal expansion of the machine or workpiece
- Fixture loosening causing parts to shift
- Material variation in the incoming stock
Corrections can include a tool offset adjustment, inserting a tool-change cycle, re-qualifying the fixture, or adjusting coolant flow.
---
Measurement Safety Reminders
- Always verify measuring instruments are calibrated and within their calibration interval before use.
- Return micrometers and gages to their cases — dropped instruments lose calibration.
- Lockout/tagout (LOTO) the machine before reaching inside to retrieve a part for measurement.
- Never measure a part while the spindle is still moving.
---
Connecting to Your Credentials
FAI and SPC knowledge appears directly in NIMS Machining Level 1 and 2 performance standards related to inspection and quality. SPC concepts also appear in Haas Certification coursework covering quality feedback and offset adjustments, and in professional Mastercam workflows where toolpath parameters are tuned based on in-process measurement data.
Key terms10
First-Article Inspection (FAI)A complete, documented measurement of every drawing-specified dimension on the first part produced from a new or revised setup, resulting in an Accept or Reject disposition.
X̄-R Chart (X-bar and Range Chart)A paired SPC control chart that tracks the subgroup mean (X̄) and subgroup range (R) over time to detect process shifts and variation changes.
SubgroupA small set of consecutive measurements taken together at one point in time, used as the basic unit of data in SPC charting.
Control LimitsStatistically calculated boundaries (UCL and LCL) placed on a control chart at approximately three standard deviations from the centerline; points outside these limits signal a special cause.
Common Cause VariationThe normal, random variation inherent in any stable process; it appears as random scatter within the control limits.
Special Cause VariationVariation from an identifiable source — such as tool wear or a bad batch of material — that produces non-random patterns or points outside control limits.
Grand Mean (X̄̄)The average of all subgroup means; it becomes the centerline of the X̄ chart and is used to calculate X̄ chart control limits.
Average Range (R̄)The average of all subgroup ranges; it forms the centerline of the R chart and is used with the A₂ constant to compute X̄ chart control limits.
Western Electric RulesA standard set of pattern-recognition rules applied to control charts to identify non-random signals that indicate a process may be out of statistical control.
PPAP (Production Part Approval Process)An automotive industry standard process for documenting that a supplier's manufacturing process can consistently produce parts meeting all engineering requirements, including a complete dimensional FAI.
Worked examples4
1. Calculate the subgroup mean (X̄) and range (R) for the following five measurements taken from one subgroup: 1.2503, 1.2498, 1.2506, 1.2501, 1.2497 inches.
- Add all five values: 1.2503 + 1.2498 + 1.2506 + 1.2501 + 1.2497 = 6.2505
- Divide by n = 5 to find X̄: 6.2505 ÷ 5 = 1.2501 inches
- Find the largest value: 1.2506
- Find the smallest value: 1.2497
- Calculate R = largest − smallest: 1.2506 − 1.2497 = 0.0009 inches
Answer: X̄ = 1.2501 in, R = 0.0009 in
2. After calculating all 25 subgroups (n = 5), you find X̄̄ = 1.2500 inches and R̄ = 0.0010 inches. Using control chart constants for n = 5 (A₂ = 0.577, D₃ = 0, D₄ = 2.114), calculate the upper and lower control limits for both the X̄ chart and the R chart.
- X̄ chart UCL: UCL_X̄ = X̄̄ + (A₂ × R̄) = 1.2500 + (0.577 × 0.0010) = 1.2500 + 0.000577 = 1.250577 inches
- X̄ chart LCL: LCL_X̄ = X̄̄ − (A₂ × R̄) = 1.2500 − 0.000577 = 1.249423 inches
- R chart UCL: UCL_R = D₄ × R̄ = 2.114 × 0.0010 = 0.002114 inches
- R chart LCL: LCL_R = D₃ × R̄ = 0 × 0.0010 = 0 inches (no lower control limit for n ≤ 6)
Answer: UCL_X̄ = 1.250577 in, LCL_X̄ = 1.249423 in, UCL_R = 0.002114 in, LCL_R = 0 in
3. On an FAI form, a dimension is called out as 0.750 ± 0.003 inches. Your measurement reads 0.7531 inches. Is the dimension in tolerance, and what is the overall disposition if this is the only dimension checked?
- Calculate the upper tolerance limit: 0.750 + 0.003 = 0.753 inches
- Calculate the lower tolerance limit: 0.750 − 0.003 = 0.747 inches
- Compare the actual measurement to the limits: 0.7531 > 0.753 (the measured value exceeds the upper limit)
- The dimension is OUT of tolerance
- Because at least one dimension is out of tolerance, the overall FAI disposition is REJECT
Answer: 0.7531 in is out of tolerance (exceeds USL of 0.753 in); overall disposition = Reject
4. On your X̄ chart, the last eight consecutive subgroup means all plot above the centerline but all remain inside the control limits. Is this process in statistical control? Which Western Electric rule applies?
- Review the Western Electric rules for non-random patterns
- Rule 4 states: eight or more consecutive points on the same side of the centerline indicate a special cause, regardless of whether they are inside the control limits
- Count the consecutive points: 8 points, all above the centerline — this satisfies Rule 4
- Conclusion: the process is NOT in statistical control even though no points are outside the control limits
Answer: The process is out of statistical control. Western Electric Rule 4 (run of 8 on one side of centerline) is violated, signaling a special cause that requires investigation.
The math in this lesson2
Learners plot subgroup means on control charts and apply run rules to judge process stability.
S-ID.1S-ID.2
In class: Measure all dimensions called out on a first-article part print (minimum ten dimensions); record results on an AS9102- or PPAP-style FAI form and determine overall part disposition. · From a provided data set of 25 subgroups of five measurements, calculate subgroup means and ranges, plot an X̄-R chart by hand, add control limits, and identify any out-of-control signals using the standard Western Electric rules. · Discuss as a class: if a control chart shows the process drifting toward the upper control limit, what measurement, tooling, or machine adjustments could bring it back in control?
Materials: Machined first-article sample part with dimensional drawing, FAI report form, SPC data worksheet, graph paper or SPC software, calculator, AIAG SPC reference guide (abridged), Machinery's Handbook.
Check: FAI report complete and disposition correct; X̄-R chart plotted with correct limits and at least one out-of-control point correctly identified; pass = FAI 100% complete and chart limits within ±5% of calculated values.
Practice check8
Answer each question, then check it. Nothing is recorded; this is your own practice.
Record this lesson on my transcript. Sign in with your key and enrollment email to record lessons (enroll free if you have no key). This page stays open to read and practise.