What it does
Creep tests at high temperature take months or years. To design a boiler tube or turbine part for 100 000 hours, engineers run shorter tests at higher temperatures and combine them with a time-temperature parameter: one number that mixes temperature and time so that all tests at the same stress fall on one master curve. This page:
- fits the Larson-Miller parameter LMP = T (C + log₁₀ t), with T in kelvin and t in hours, as a straight line, quadratic or cubic in log₁₀ σ. C = 20 is the usual convention; the page also finds the C that fits your own data best.
- fits two other classic parameters for comparison: Sherby-Dorn P = log₁₀ t − Q/(2.303 RT), with an activation energy Q, and Manson-Haferd P = (T − Ta)/(log₁₀ t − log₁₀ ta), whose constants are the point where the lines of equal stress meet. You can type the constants or let the page fit them.
- predicts the rupture life at any temperature and stress and the stress for a design life, by all three methods side by side.
- fits Norton’s creep law ε̇ = A σⁿ exp(−Q/RT) to minimum (steady-state) creep rates, giving the stress exponent n and, with tests at two or more temperatures, the activation energy Q.
Each fit reports its scatter as the standard deviation of log₁₀ t between the tests and the curve, so you can see which parameter describes your data best.
How to use it
- Paste temperature, stress and rupture time in hours, or open a CSV or Excel file. Choose °C or K.
- Pick the polynomial. A quadratic suits most data; use a straight line for few tests.
- Read the master curve, the best-fit C and the comparison table. The plot shows stress against LMP ÷ 1000, coloured by temperature: points from different temperatures should fall on one curve.
- Type a temperature, stress and design life to get predictions.
- For creep rates, paste stress, temperature and minimum creep rate.
Worked example
The example has 12 made-up tests at 600, 650 and 700 °C. With C = 20 and a quadratic, the master curve is LMP = 26 056 + 1235.8 x − 1383.7 x² (x = log₁₀ σ) with a scatter of 0.077 in log t. The best-fit C is 19.50 (scatter 0.076). Sherby-Dorn fits Q = 417 kJ/mol (scatter 0.077) and Manson-Haferd Ta = 500.8 K, log ta = 14.89 (scatter 0.075).
At 650 °C and 120 MPa, Larson-Miller predicts 33 784 h, Sherby-Dorn 32 579 h and Manson-Haferd 32 647 h. For a design life of 100 000 h at 600 °C, Larson-Miller allows 178.1 MPa.
The made-up creep rates give n = 5.07 and Q = 284 kJ/mol (they were drawn with n = 5 and Q = 280 kJ/mol), typical of dislocation creep.
Tips
- Use only tests that failed by creep (not by oxidation or a test fault), and one heat or product form at a time.
- Extrapolating more than about three times beyond your longest test is risky. Compare the three methods: if they disagree, be cautious.
- Design codes usually take the stress for rupture in 100 000 h, divide by a safety factor, and use the lower bound of the scatter, not the mean.
- n near 1 means diffusion creep; 3 to 8, dislocation creep. A change in n with stress means a change in mechanism, so fit each regime separately.
FAQ
Why is C about 20? Larson and Miller found that, for many steels, lines of equal stress on a plot of log t against 1/T meet near log t = −20 at 1/T = 0. Fitting C to your own data is better when you have enough tests.
Do I need hours? Yes, for the parameters: the constants C and log ta depend on the time unit. For Norton’s law any rate unit works.
Is my data uploaded? No. Everything runs on this page.
Where the numbers come from
- The Larson-Miller, Sherby-Dorn and Manson-Haferd parameters and Norton's law are the textbook forms (Dieter, Mechanical Metallurgy, chapter 13; Dowling, Mechanical Behavior of Materials, chapter 15), written by fibub and checked in the tests against NumPy (polyfit, lstsq) and SciPy (minimize_scalar, Nelder-Mead, brentq).
- The example creep-rupture and creep-rate data are made up by the page from a chosen master curve and Norton law with random scatter; they are not data for any real alloy.
- Your data is read on the page only. It is never uploaded or stored.
Standard scientific values, such as constants, atomic weights and melting points, are facts found in every textbook; we name where ours were checked. If you are an author or publisher and want a value, a credit or a table changed or removed, tell us and we will act quickly.
Last updated 2026-11-09.