Worst-Case vs. RSS Tolerance Analysis: When to Use Each Method

Choosing between worst-case (WC) and root sum of squares (RSS) analysis affects your part costs, assembly yield, and safety margins. This guide explains both methods and helps you decide which is right for your application.

The Core Difference

Worst-case analysis asks: "What happens if every single dimension lands at its worst limit?" This is physically possible but statistically unlikely. If your WC stack passes, 100% of assemblies will work.

RSS analysis asks: "What's the probable range of assembled dimensions?" It uses statistical mathematics to predict that most parts will be closer to nominal, so tolerances partially cancel out. RSS gives tighter predicted ranges but accepts some small percentage of assemblies may fall outside.

When to Use Worst-Case Analysis

Choose worst-case when:

• Safety is critical — Aircraft, medical devices, automotive safety systems
• Low production volume — Prototype or custom builds where statistical averaging doesn't apply
• 100% interchangeability required — Field-replaceable parts with no sorting or matching
• Process capability is unknown — New suppliers or processes without Cpk data
• Assembly cannot be inspected — Sealed units or inaccessible interfaces

When to Use RSS Analysis

Choose RSS when:

• High production volume — Thousands of units where statistics apply
• Processes are in control — Demonstrated Cpk ≥ 1.33 on contributing dimensions
• Cost reduction needed — WC tolerances would require expensive precision
• Some fallout is acceptable — You can inspect and scrap the rare failures
• Parts are from multiple suppliers — Variation sources are independent

The Math Behind Each Method

Worst-Case Calculation

Simply add up all the tolerance magnitudes:

RSS Calculation

Square each tolerance, sum them, take the square root:

Notice RSS (±0.137) is about 40% smaller than WC (±0.23). This difference grows with more dimensions in the stack.

Comparing Results: A Real Example

Consider a 6-dimension stack with ±0.05mm on each dimension:

RSS predicts 59% less variation. If your assembly can tolerate ±0.20mm but WC gives ±0.30mm, RSS analysis might justify using cheaper, looser tolerances.

Running Both: The Engineering Approach

Smart engineers run both analyses and compare:

1. Calculate WC first — If it passes, you're done. Design is robust.
2. If WC fails, calculate RSS — See if statistical approach saves the design.
3. Document both — Show reviewers you considered the trade-off.
4. Make an explicit decision — State which method applies and why.

Common Mistakes

Industry Standards Reference

These standards discuss tolerance analysis methods:

• ASME Y14.5 — GD&T fundamentals
• ISO 8015 — Fundamental tolerancing principle
• SAE AS9103 — Variation management (aerospace)
• AIAG Core Tools — Measurement system analysis

Summary: Decision Matrix