Precision Meets Simplicity with Our Tolerance Analysis Tool
Evaluate stacks directly in your browser using RSS and Worst-Case (Max/Min) analysis. Simply enter the loop direction, nominal values, bilateral or asymmetric tolerances, and sigma capability, and then see immediate assembly estimates along with pass/fail results against your specifications.
If you need deeper insight, you can upgrade to a Premium Membership. With this upgrade, you unlock Monte Carlo simulation, which provides true distribution and yield analysis that is reproducible via a seed. In addition, you gain access to one-click, audit-ready PDF reports that include full input and results tables. Start for free, and upgrade anytime to expand your analysis capabilities.
Using the calculator
Tolerance Analysis Calculator FAQ
Understand your inputs, compare analysis methods, and interpret assembly variation with these answers to common tolerance stack-up questions.
What is tolerance stack-up analysis?
Tolerance stack-up analysis estimates how variation in individual part dimensions affects an assembly dimension, such as a clearance, gap, or overall length. Define the dimension loop, enter each nominal value and tolerance, and use the results to assess whether the assembly can meet its dimensional requirements.
How do worst-case, RSS, and Monte Carlo analysis differ?
Worst-case analysis combines the dimensional limits to find the largest possible stack-up range within the entered tolerances. Root sum square (RSS) estimates statistical variation using the assumed standard deviations. Monte Carlo simulates many assemblies to estimate the output distribution. Statistical results depend on the model assumptions and do not guarantee that every assembly will fit.
How do I enter dimensions, tolerances, and loop directions?
Add one row for each dimension in your loop and use the same unit throughout, such as millimeters or inches. For a dimension of 10 ± 0.1, enter 10 as the nominal dimension and 0.1 as the bilateral tolerance, not the full 0.2 range. Choose + when the dimension adds to the target dimension and − when it subtracts. Source Part and Description help identify each row.
Can I use unilateral or unequal bilateral tolerances?
The calculator accepts symmetric bilateral tolerances. Convert unequal limits to their midpoint and half-range before entering them. For example, 10 +0.2/−0.0 has limits of 10.0 and 10.2, so enter 10.1 ± 0.1. This preserves the dimensional limits; it does not establish the actual process mean or distribution for a statistical analysis.
What does the sigma factor mean?
The sigma factor k relates the entered half-tolerance T to the assumed standard deviation: σ = T / k. Leaving it blank uses k = 3. For example, ±0.12 with k = 3 implies σ = 0.04. Change k only when supported by process information; a drawing tolerance alone does not prove a manufacturing process has that capability.
How should I interpret specification limits, scrap percentage, and Cpk?
Lower and upper specification limits define the acceptable range for the assembled result. Enable them and enter both limits to evaluate estimated out-of-spec percentage and capability. Cpk compares the modeled mean and spread with the nearer specification limit. These are model-based estimates, not a substitute for capability studies using measured data from a stable process.
How many Monte Carlo samples should I use?
The default is 50,000 samples. More samples generally reduce simulation noise but take longer to calculate; use the seed field to reproduce a run with the same inputs and settings. A simulation showing zero out-of-spec samples does not prove zero risk, particularly for rare events. Monte Carlo availability depends on the calculator edition.
Can I save an analysis and generate a PDF report?
Use Save to download a JSON project file, then Load to reopen it in the calculator. The project file preserves your inputs and analysis settings so you can continue later. Add your project information, objective, conclusion, and recommendation under Report Details, then select Generate Report to export a PDF for review or documentation.
What assumptions and limitations should I check?
This is a one-dimensional, linear dimension-loop model. The statistical methods assume independent dimensional variations and use normal-distribution assumptions; this Monte Carlo implementation samples normal variations. Correlated dimensions, non-normal processes, geometric tolerances, deformation, and complex assembly constraints may require a different model. Validate the inputs and assumptions against the actual assembly and manufacturing process.
Where can I find a worked tolerance analysis example?
Follow our RSS and worst-case tolerance stack-up example. It shows how to find the clearance between two blocks and a pocket, create the dimension loop, convert a unilateral tolerance, enter the values, and interpret the results. Use it as a step-by-step guide when setting up your own analysis.
