How to Prepare a Tolerance Stack-Up for Design Review
A stack-up that passes internal review is one thing. A stack-up that survives design review with customers, quality, or certification bodies is another. Here is how to prepare one that holds up.
The Goal
Make the reviewer's job easy. A good stack-up answers questions before they are asked: What did you assume? What method did you use? What drives the variation? What happens if tolerances shift?
1. Document Your Assumptions
Every stack-up rests on assumptions. Undocumented assumptions become arguments in design review. Common assumptions that need explicit documentation:
- Temperature — Are parts at room temperature? Operating temperature? What about thermal expansion?
- Assembly sequence — Does order matter? Is there a specific datum flow?
- Coaxiality/alignment — Are features perfectly aligned or offset?
- Statistical basis — 3σ? 4σ? 6σ? What Cpk are you assuming?
- Tolerance interpretation — How did you convert GD&T to bilateral?
- Excluded factors — Assembly variation, wear, fastener play?
Use a checklist format. Check the ones you verified, leave unchecked the ones you assumed. Reviewers know where to probe.
Example Assumptions Checklist
- ☑ All parts at 20°C ± 2°C
- ☑ Assembly per SOP-1234
- ☐ No thermal expansion considered
- ☑ Position tolerances converted at MMC
- ☐ Fastener play excluded (worst-case seated)
2. Write Method Notes
Method notes explain the “why” behind your approach. Not everyone in the room knows your reasoning. Document:
- Why this analysis method? — Why WC, RSS, or both?
- Dimension chain logic — How does the chain close? What is the gap representing?
- Sign conventions — What does positive/negative mean in your stack?
- Sources — Where did nominal/tolerance values come from? Drawing rev? Measurement data?
Keep notes concise but complete. A few sentences is usually enough.
3. Show Both WC and RSS (Usually)
Most design reviews expect to see both analyses. The question is not “which one?” but “what do they each tell us?”
Worst-Case
Shows the absolute boundaries. If WC passes, you guarantee 100% assemblability. If WC fails, you need to justify why RSS is acceptable.
RSS (Statistical)
Shows the expected variation for production volume. Tighter than WC. Requires statistical process control to be valid.
When to show only one:
- WC only — Safety-critical, low volume, prototype, or process capability unknown
- RSS only — High volume, proven Cpk, and WC would require unreasonable precision
When in doubt, show both. Let the reviewer decide which drives the design.
4. Use Sensitivity Analysis
Sensitivity answers the inevitable question: “What should we tighten?”
What Sensitivity Shows
The percentage each dimension contributes to RSS variation. A dimension at 45% contributes nearly half your total variation. Tightening it has far more impact than tightening a 5% contributor.
In design review, sensitivity lets you have informed discussions:
- “The housing bore drives 52% of variation. We should focus there.”
- “The shaft tolerance contributes 8%—tightening it won't meaningfully improve the stack.”
- “If we can hold ±0.02 on the plate instead of ±0.05, RSS drops below target.”
Always include a sensitivity breakdown. It transforms stack-ups from “pass/fail” into actionable engineering data.
5. Present Professionally
Format matters. A stack-up buried in a messy spreadsheet undermines confidence. A clean, consistent report builds trust.
What a Design Review Report Needs
- Header — Project name, revision, date, analyst
- Assembly target — Min/max acceptable gap or clearance
- Dimension table — Name, nominal, tolerance, direction, notes
- Results — WC min/max, RSS min/max, pass/fail status
- Sensitivity — Top contributors ranked
- Assumptions — Checklist or bullet list
- Method notes — Brief explanation of approach
PDF is the standard format for formal reviews. It preserves formatting, is easy to archive, and can be attached to design records.
6. Anticipate Questions
Prepare for the questions reviewers will ask:
- “What if the tolerance is wider?” — Run a sensitivity case with looser tolerances
- “Is this 3σ or 6σ?” — Document in assumptions
- “Where did these tolerances come from?” — Reference drawing revision
- “What about thermal effects?” — Either include them or note they are excluded and why
- “Can we tighten X instead of Y?” — Sensitivity analysis answers this
7. Version Control
Stack-ups evolve. Track changes properly:
- Include revision number in the report header
- Note what changed from the previous version
- Archive superseded versions (don't overwrite)
- Export to JSON for backup and portability
Design Review Checklist
Before presenting your stack-up, verify:
- Assumptions documented (temperature, assembly, statistical basis)
- Method notes explain dimension chain and approach
- Both WC and RSS results shown (or justification for one)
- Sensitivity analysis identifies top contributors
- Report formatted professionally (PDF preferred)
- Target min/max clearly stated with pass/fail
- Dimension sources traceable to drawing/data
Summary
A design-review-ready stack-up is more than correct math. It is documented assumptions, clear method notes, both WC and RSS results, sensitivity analysis, and professional formatting. Do the work upfront, and the review becomes a discussion of engineering trade-offs instead of an interrogation of your methods.
Start your stack-up in TolReport → — assumptions checklist, sensitivity analysis, and PDF export built in.