Online Weibull analysis and reliability calculator
Fit a Weibull distribution to failure data and get the shape parameter β, the characteristic life η, the B1, B10 and B50 life, the mean time to failure and the reliability at any mission time. The Weibull probability plot shows whether the model fits, and the reliability curve shows how many units survive over time.
Reliability engineers, quality engineers and product developers use it for life tests, field return analysis, warranty forecasts and maintenance planning.
How the Weibull Analysis works
Paste the times to failure in hours, cycles, kilometres or days. The two-parameter Weibull is fitted by maximum likelihood, the same method as SciPy weibull_min, and the plot uses Bernard median ranks. β below 1 points to early failures, about 1 to random failures and above 1 to wear-out.
Enter a mission time to read the reliability at that point, for example the share of units still working at the end of the warranty period.
When to use it
Use Weibull analysis after a life test, when field returns accumulate, when a customer asks for a B10 life, or when you decide whether preventive replacement makes sense. A shape parameter above 1 means replacement before the B10 life prevents failures; a value around 1 means failures are random and preventive replacement only adds cost.
With fewer than about six failures the parameters are very uncertain, so treat early results as indicative.
PDF report in one click
The PDF report contains your header with component and test, the failure mode interpretation, β and η, the life table, the Weibull plot and the reliability curve.
Private: everything runs in your browser
The Weibull fit runs locally in your browser; failure data stays on your device and nothing is uploaded.
How to run a Weibull analysis
- Collect failure times from a life test or field returns, in hours, cycles or kilometres.
- Paste them into the column and choose the unit.
- Read β to understand the failure mode and B10 for the life requirement.
- Enter a mission time to see the reliability at that point.
Use the result to plan preventive maintenance (wear-out, β > 1), to check a supplier's life test against the requirement, or to size a zero-failure test.
Frequently asked questions
›What do β and η mean?
β (shape) tells the failure mode: below 1 early failures, about 1 random failures, above 1 wear-out. η (scale or characteristic life) is the time by which 63.2 % of the units have failed.
›What is the B10 life?
The time by which 10 % of the units are expected to fail. It is a common reliability requirement for bearings, components and products.
›Can I include units that did not fail?
This version fits complete failure data (all units failed). Suspended (censored) units will be supported in a later version; until then, analyse only the failures and treat the result as conservative.
›How accurate is the fit?
The maximum likelihood fit is validated against SciPy weibull_min. With fewer than about 6 failures the estimates are very uncertain.