Tolerance Stack-Up Loop Diagram

A loop diagram is the visual foundation of any tolerance stack-up. It shows exactly which dimensions contribute to your critical measurement and in which direction. Getting the loop right means getting the analysis right. This guide walks through how to construct one from scratch.

What Is a Loop Diagram?

A loop diagram (also called a vector loop or dimension chain) traces a continuous path through your assembly, connecting all the dimensions that affect your critical measurement. The path forms a closed loop—starting and ending at the same point.

Think of it as walking along the edges of your parts from one side of the gap to the other, then jumping across the gap to close the loop. Every step you take is a dimension in your stack.

Start and End at the Gap

The "gap of interest" is what you're trying to calculate—the clearance, fit, or critical distance that must fall within spec. Your loop starts on one side of this gap and ends on the other side.

Example: For a clevis pin assembly, if the gap is the clearance between the pin shoulder and the clevis ear, start your loop at the left face of the gap and trace dimensions until you reach the right face of the gap.

The Signed Direction Convention

As you trace the loop, each dimension gets a sign based on its direction:

• Positive (+) — Dimension goes in the "positive" direction (toward closing the gap)
• Negative (−) — Dimension goes in the "negative" direction (away from closing the gap)

The convention you choose doesn't matter as long as you're consistent. A common approach:

How Signs Affect the Math

When you increase a positive dimension, the gap increases. When you increase a negative dimension, the gap decreases. This is crucial for understanding which tolerances hurt you and which help you.

Closing the Loop

A properly constructed loop must close—meaning you return to your starting point after accounting for all dimensions. The "closing dimension" is usually the gap you're solving for.

Mathematically:

Or equivalently:

If your loop doesn't close (you can't get back to the start), you've missed a dimension or taken a wrong path.

Worked Example: Simple Bracket Assembly

A bracket holds a shaft with a retaining ring. The gap of interest is the axial clearance between the ring and the bracket face.

Assembly Sketch (Text Representation)

    ┌─────────────────────────────────┐
    │         BRACKET                 │
    │    ┌─────────────────────┐      │
    │    │                     │      │
    │    │      ┌───────┐      │      │
    │    │      │ SHAFT │      │      │
    │    │      └───────┘      │      │
    │    │         ├──┤        │      │
    │    │         RING        │      │
    │    │                     │      │
    │    │◄──────GAP──────────►│      │
    │    └─────────────────────┘      │
    └─────────────────────────────────┘

    Dimension A: Bracket pocket depth (left wall to right wall)
    Dimension B: Shaft shoulder position (left face to ring groove)
    Dimension C: Ring thickness
    Gap: Clearance between ring and bracket wall

Building the Loop

Start at the left side of the gap (ring face) and trace to the right side (bracket wall):

The Loop Equation

With Numbers

Common Mistakes

1. Double-Counting Dimensions

If you pass through the same feature twice (e.g., going around a part instead of through it), you'll count that dimension twice. The loop should be the shortest path through the assembly.

2. Wrong Sign Assignment

The most common error. Ask yourself: "If this dimension gets bigger, does the gap get bigger or smaller?"

• If bigger dimension → bigger gap: same sign as gap direction
• If bigger dimension → smaller gap: opposite sign

3. Missing Dimensions

Every surface or feature you cross needs a dimension. Common omissions:

• Washer or shim thicknesses
• Coating or plating thickness
• Snap ring groove positions
• Chamfers or radii that affect contact points

4. Confusing the Gap Sign

Some engineers put the gap on the "wrong side" of the equation. Remember: the gap is just another dimension in the loop. Its sign depends on which direction you cross it.

5. Not Validating with Nominal

Before running WC or RSS, check that nominal dimensions give a sensible nominal gap. If your nominal gap is negative when it should be positive, you have a sign error.

Tips for Complex Loops

• Sketch it first — Draw arrows on your assembly drawing showing the loop path before building the spreadsheet.
• Number each dimension — Assign D1, D2, D3... and mark them on the sketch so nothing gets lost.
• Color-code signs — Use green for positive, red for negative dimensions on your sketch.
• Break into sub-loops — For assemblies with multiple gaps, trace each loop separately. Some dimensions may appear in multiple loops.
• Sanity-check with extreme cases — Mentally set all positive dimensions to max and negatives to min. Does the result make physical sense?

Key Takeaways

1. Start and end at the gap — The loop connects both sides of your critical dimension
2. Assign signs by direction — Pick a convention and stick to it consistently
3. Close the loop — If you can't return to start, you missed something
4. Each dimension appears once — Double-counting inflates your tolerance
5. Validate with nominals first — Catch sign errors before the full analysis