Manufacturing Engineering Calculator

Process Capability (Cp/Cpk) Calculator

Calculate statistical capability indices (Cp, Cpk) to evaluate how well a manufacturing process performs relative to its defined specification limits.

Process Distribution

Normal distribution relative to specification limits

Input Parameters

Enter your statistical process data.

Upper Spec Limit (USL)

Maximum acceptable limit.

Lower Spec Limit (LSL)

Minimum acceptable limit.

Process Mean (μ)

Average value of the measured data.

Standard Deviation (σ)

Process standard deviation.

Target Cpk (Limit)

Industry standard is typically 1.33 or 1.67.

Quality Tip

A process with a high Cp but low Cpk is capable but not centered. Adjusting the machine offset (mean shift) is usually easier than reducing variance.

Process Capability Index (Cpk)

0.667

Cp (Potential)

1.111

Cpl (Lower)

1.556

Cpu (Upper)

0.667

NOT CAPABLE — Process fails to meet the target

Calculated Cpk is 0.667 compared with the target limit of 1.33.

Governing Formulas

Cp = (USL - LSL) / 6σCpk = min( (USL - μ) / 3σ , (μ - LSL) / 3σ )

USLunits

Upper Specification Limit

LSLunits

Lower Specification Limit

μ (Mean)units

Average of the sample data

σ (StdDev)units

Process standard deviation

Calculation Assumptions

Data is normally distributed
Process is strictly stable/in-control
Representative sample size (n > 30)
Independent observations

Quality Control Code

Reuse the SPC calculation in your own data analysis workflow.

Python

def process_capability(usl, lsl, mean, std_dev):
    """
    Calculate Process Capability Indices (Cp, Cpk).
    """
    if std_dev <= 0:
        raise ValueError("Standard deviation must be > 0")
    if lsl >= usl:
        raise ValueError("USL must be greater than LSL")

    cp = (usl - lsl) / (6 * std_dev)
    cpu = (usl - mean) / (3 * std_dev)
    cpl = (mean - lsl) / (3 * std_dev)
    cpk = min(cpu, cpl)
    
    return cp, cpk, cpu, cpl

# Example Parameters
USL = 10.5
LSL = 9.5
Mean = 10.2
StdDev = 0.15

cp, cpk, cpu, cpl = process_capability(USL, LSL, Mean, StdDev)

print(f"Cp: {cp:.3f} | Cpk: {cpk:.3f}")

MATLAB

function [Cp, Cpk] = process_capability(USL, LSL, Mean, StdDev)
    % Calculate Process Capability Indices
    
    if StdDev <= 0
        error('Standard deviation must be > 0');
    end
    if LSL >= USL
        error('USL must be greater than LSL');
    end

    Cp = (USL - LSL) / (6 * StdDev);
    Cpu = (USL - Mean) / (3 * StdDev);
    Cpl = (Mean - LSL) / (3 * StdDev);
    Cpk = min(Cpu, Cpl);
end

% Example Parameters
USL = 10.5;
LSL = 9.5;
Mean = 10.2;
StdDev = 0.15;

[Cp, Cpk] = process_capability(USL, LSL, Mean, StdDev);
fprintf('Cp: %.3f | Cpk: %.3f\n', Cp, Cpk);

Excel Formula (Cpk)

=MIN((USL-Mean)/(3*StdDev), (Mean-LSL)/(3*StdDev))

Example Calculation

Consider a CNC turned shaft with a diameter specification of 10.0 ± 0.5 mm (USL = 10.5, LSL = 9.5). If the production sample yields a mean of 10.2 mm and a standard deviation of 0.15 mm:

Cpu = (10.5 - 10.2) / (3 × 0.15) = 0.667

Cpl = (10.2 - 9.5) / (3 × 0.15) = 1.556

Cpk = min(0.667, 1.556) = 0.667

Result: The Cpk is 0.667, which is far below the standard 1.33 target. The process is significantly shifted towards the upper limit (USL).

Technical Explanation: Process Capability (SPC)

Process capability analysis quantifies how well a stable manufacturing process satisfies its engineering specification limits. Rather than measuring defects after they occur, capability indices compare the natural statistical variation of the process — its spread and centering — against the tolerance window defined by the design engineer (USL and LSL). The two most widely used indices, Cp and Cpk, form the core of Statistical Process Control (SPC) and are prerequisites for PPAP submissions, supplier qualification, and Six Sigma projects.

Both indices assume the process output follows an approximately normal distribution, which is reasonable for most machining, molding, and assembly measurements once special-cause variation has been eliminated. The indices are dimensionless ratios, meaning they apply equally whether the characteristic is a shaft diameter in millimeters or a coating thickness in microns.

Governing Formulas

Cp (Potential Capability) measures only the spread of the process relative to the tolerance width:

Cp = (USL − LSL) / (6σ)

where USL and LSL are the upper and lower specification limits, and σ is the process standard deviation estimated from a rational subgroup of samples (typically n ≥ 30). The denominator 6σ represents the full natural process width — the interval μ ± 3σ that encloses 99.73% of a normal distribution. The formula originates from the empirical rule of the normal distribution: if the specification window is wider than six sigmas, nearly all output falls within tolerance.

Cpk (Actual Capability) additionally accounts for process centering. It is defined as the minimum of two one-sided indices:

Cpu = (USL − μ) / (3σ) | Cpl = (μ − LSL) / (3σ)

Cpk = min(Cpu, Cpl)

Each one-sided index divides the distance from the process mean μ to the nearest specification limit by half the process width (3σ). Taking the minimum ensures the result reflects the worst side of the distribution — the side where defects will actually occur first.

How to Interpret Cpk Values

Cpk < 1.0: The process is incapable. The 3σ width on the nearest side exceeds the available margin, so scrap or rework is actively occurring. A negative Cpk means the mean lies entirely outside the limits — a severe, non-conforming condition.
Cpk = 1.0: Barely capable. Exactly 99.73% of output is in spec (3 Sigma level), leaving zero margin for tool wear, thermal drift, or material variation. Any shift causes defects.
Cpk = 1.33: The standard minimum for most industries, corresponding to a 4 Sigma level with a modest safety margin of 33%.
Cpk = 1.67: Frequently required for safety-critical characteristics in automotive and medical applications.
Cpk = 2.0: Six Sigma capability. Accounting for the conventional 1.5-sigma long-term drift, the expected defect rate is 3.4 parts per million.

Input Parameters

USL / LSL: The engineering tolerance limits from the drawing or control plan — not inspection limits. Entering incorrectly tightened "shop limits" inflates capability falsely.
Mean (μ): The arithmetic average of your sample. Off-center means are the most common reason a capable machine produces bad parts.
Standard Deviation (σ): Estimated from within-subgroup variation (e.g., the average range method or pooled standard deviation), not the total variation across shifts, which would mix special causes into the estimate.

Cp vs. Cpk: Reading the Difference

When Cp = Cpk, the process is perfectly centered. When Cp > Cpk, the variation is small enough to fit the tolerance, but the mean is shifted toward one limit — the improvement path is simply recentering via offset adjustment, which is fast and inexpensive. When both indices are low, the fundamental process variation is too large; recentering will never help, and you must attack σ through tooling, thermal control, or process redesign. This diagnostic split is precisely why engineers report both numbers together.

Assumptions and Limitations

The process must be statistically in control (stable, predictable mean and spread) before capability indices are meaningful. An out-of-control process has no definable σ.
Indices assume approximate normality. For skewed data (e.g., flatness, surface roughness, contamination counts), non-normal capability methods (Weibull, Box-Cox) should be used instead.
Sample size matters: Cpk computed from a handful of parts carries wide confidence intervals and can overstate true capability substantially.
Cp and Cpk are short-term capability measures. Long-term performance (Pp/Ppk) will typically be lower due to drift between runs, operators, and raw material batches.
High capability does not guarantee zero defects: a single out-of-control event, a mis-set offset, or a wrong material lot can produce scrap despite excellent historical indices.

Practical Engineering Use

Capability studies appear at every stage of manufacturing quality: as initial process studies in PPAP and supplier approval, as ongoing control-plan checkpoints on SPC charts, in machine and fixture qualification (run-off and capability runs), and as the baseline metric in Six Sigma DMAIC projects. Because Cpk condenses mean position and variation into one comparable number, it allows quality engineers to rank processes, prioritize improvement resources, and set evidence-based acceptance criteria for production sign-off.

Continue Calculating

calculators

Calculators

Quickly calculate and analyze parameters for your calculators requirements.

Open

calculators

Civil

Quickly calculate and analyze parameters for your civil requirements.

Open

calculators

Beam Bending Moment

Quickly calculate and analyze parameters for your beam bending moment requirements.

Open

Real-World Engineering Cases

Tool Wear in High-Volume CNC Machining

A high-volume supplier of automotive transmission shafts maintained a Cp of 1.80, proving their modern lathes were highly precise. However, the quality manager neglected tool wear offsets. Over a single shift, the mean diameter drifted drastically towards the Upper Spec Limit (USL).

Engineering Lesson

The batch resulted in a Cpk of 0.85 despite the excellent Cp. Precise machines (low standard deviation) still require strict offset monitoring (centering the mean) to maintain high Cpk.

Injection Molding Cavity Variations

A plastic injection molding process was set up with the mean perfectly centered between limits (Cp = Cpk). However, due to uneven cooling lines in a multi-cavity mold, the standard deviation across parts was massive.

Engineering Lesson

Both Cp and Cpk dropped below 1.0. When Cp is poor, adjusting the machine parameters (the mean) will not save you. You must fix the fundamental process variation (in this case, thermal management) to reduce standard deviation.

Frequently Asked Questions

What is the difference between Cp and Cpk?

Cp measures the potential capability of a process if it were perfectly centered. Cpk measures the actual capability, taking into account where the process mean is currently located.

What is considered a good Cpk value?

In most manufacturing, a Cpk of 1.33 is the minimum standard. For aerospace or critical safety systems, a Cpk of 1.67 or 2.00 (Six Sigma) is heavily preferred.

Why is my Cpk lower than my Cp?

This happens when your process is off-center (process shift). The variation is tight, but the average value is sitting too close to one of the specification limits.

Can Cpk be negative?

Yes. If the process mean falls entirely outside of the specification limits, the Cpk value will be negative, indicating a severely non-conforming process.