Sign in to save

Bookmark this page so you can find it later.

Sign in to save

Bookmark this page so you can find it later.

Tolerance stack-up analysis predicts how individual part variations combine to affect an assembly dimension such as a gap, clearance, preload, or interference fit. It matters because every manufactured part has allowable variation, and small deviations can add together to create a failed fit even when each part is within its own tolerance. Engineers use stack-up analysis to decide whether a design can be built reliably, inspected efficiently, and produced at a reasonable cost.

A stack-up begins by defining a dimension chain from one functional surface to another, then assigning each contributing dimension a sign based on whether it increases or decreases the final result. Worst-case analysis assumes all dimensions hit their most unfavorable limits at the same time, while statistical RSS analysis assumes independent random variations combine by root-sum-square. Tolerance allocation works in the reverse direction by dividing an allowable assembly variation among parts based on function, process capability, cost, and risk.

Key Facts

  • Assembly result: R = d1 + d2 - d3 + d4, using signs from the dimension chain.
  • Worst-case tolerance: T_wc = T1 + T2 + T3 + ... + Tn.
  • Worst-case limits: R_min = R_nominal - T_wc and R_max = R_nominal + T_wc when T values are bilateral half-widths.
  • Statistical RSS tolerance: T_rss = sqrt(T1^2 + T2^2 + T3^2 + ... + Tn^2).
  • For independent normal dimensions, standard deviations combine as sigma_R = sqrt(sigma_1^2 + sigma_2^2 + ... + sigma_n^2).
  • If a dimension has tolerance ±T and represents a 3 sigma process, then sigma = T / 3.

Vocabulary

Tolerance stack-up
A calculation that determines how part dimension variations combine to affect a final assembly dimension.
Dimension chain
The ordered path of dimensions that connects the two surfaces defining the functional assembly requirement.
Worst-case analysis
A method that assumes every part dimension is at its extreme limit in the direction that creates the largest or smallest assembly result.
RSS analysis
A statistical method that combines independent tolerances using the square root of the sum of their squares.
Tolerance allocation
The process of assigning allowable variation to individual part dimensions so the assembly requirement is met.

Common Mistakes to Avoid

  • Adding all nominal dimensions without signs, because some dimensions increase the gap while others reduce it.
  • Using RSS for dependent dimensions, because shared tooling, common datums, or process shifts can make variations correlated rather than independent.
  • Treating total tolerance width as the plus-minus value, because a dimension listed as 20.00 ± 0.10 has a half-width of 0.10 and a full width of 0.20.
  • Allocating equal tolerances to every part automatically, because critical surfaces, manufacturing processes, and cost sensitivity often require unequal tolerances.

Practice Questions

  1. 1 A gap is defined by G = A - B - C. If A = 50.00 ± 0.20 mm, B = 18.00 ± 0.10 mm, and C = 31.50 ± 0.15 mm, find the nominal gap and the worst-case minimum and maximum gap.
  2. 2 Four independent dimensions contribute to a critical length with bilateral tolerances ±0.05 mm, ±0.08 mm, ±0.04 mm, and ±0.06 mm. Calculate the worst-case tolerance and the RSS tolerance.
  3. 3 An assembly has a tight functional gap requirement, but one spacer is produced by a low-cost process with poor repeatability. Explain how tolerance allocation could be changed to improve yield without making every part more expensive.