GOLD STANDARD · POLYMER SCIENCE · HOW X WORKS
Find the decision hidden inside the model
Time–temperature superposition turns several short viscoelastic tests at different temperatures into a longer apparent time or frequency window. The construction is powerful only when the material is thermorheologically simple, the shift convention is explicit and the master curve is not mistaken for an unlimited lifetime prediction.
This guide connects the Gold Standard of Polymer Science, Gold Standard of Education, and How X Works hub.
Chapter contents and reading routes
Choose the route closest to your present task. Every chapter remains expanded, so the article also reads naturally from start to finish.
Define superposition and axes
Build a shift-factor ledger
Fit and interpret WLF
Test transfer and failure modes
Use sources and connected routes
CHAPTER 1 OF 12 · DEFINE SUPERPOSITION
1. Temperature changes the clock of viscoelastic response
A polymer can appear stiff during a fast test and softer during a slow one because molecular rearrangements need time. Raising temperature often accelerates those processes, so a response measured at one temperature may resemble a response measured over a different time scale at another.
Time–temperature superposition asks whether curves at several temperatures have the same shape and differ mainly by a horizontal shift along log time or log frequency. It is a testable structural claim, not a licence to slide any curves until they look smooth.
CHAPTER 2 OF 12 · DEFINE SUPERPOSITION
2. Choose the reference and shift convention first
Let aT be the horizontal shift factor from temperature T to reference Tref. One common frequency-domain convention forms reduced frequency fR = aT f; other texts use the reciprocal. The sign of log aT therefore depends on the declared convention.
Write the mapping beside every table and plot. A physically correct shift can look numerically reversed when conventions are mixed. State whether a vertical modulus or density correction is also used and whether the reported response is storage modulus, loss modulus, compliance or another quantity.
CHAPTER 3 OF 12 · BUILD THE CURVE
3. Build a small master-curve ledger
Take three illustrative loss-modulus peaks measured at 10°C, 20°C and 30°C. Choose 20°C as reference. Suppose aligning the 10°C curve requires aT = 100 and aligning the 30°C curve requires aT = 0.01 under fR = aT f.
A 1 Hz point at 10°C maps to 100 Hz reduced frequency; a 1 Hz point at 30°C maps to 0.01 Hz. These are constructed values for arithmetic, not material constants. The aligned overlap region is the first evidence check.
| Temperature | aT | Measured f | Reduced f = aT f |
|---|---|---|---|
| 10°C | 100 | 1 Hz | 100 Hz |
| 20°C | 1 | 1 Hz | 1 Hz |
| 30°C | 0.01 | 1 Hz | 0.01 Hz |
CHAPTER 4 OF 12 · BUILD THE CURVE
4. A master curve is assembled from overlaps
Shift neighbouring curves first, using a reproducible objective over their overlap rather than anchoring distant temperatures independently. Check whether storage and loss responses imply compatible shifts. Preserve the raw curves, chosen window, weights and optimisation criterion.
A visually smooth join can conceal sparse overlap or a changing curve shape. Plot residual mismatch and the uncertainty in each shift factor. If different observables demand different aT values, the material may not be thermorheologically simple over that range.
CHAPTER 5 OF 12 · BUILD THE CURVE
5. WLF relates shift factor to temperature near Tg
The Williams–Landel–Ferry relation is commonly written log10 aT = −C1(T−Tref)/(C2+T−Tref). C1 and C2 are empirical constants tied to the material, reference temperature, units and fitted range.
IUPAC notes that WLF is usually applied near the glass-transition region and that its constants have a limited temperature range. Quoting “universal” constants without their convention or material context can create a precise-looking but invalid shift.
CHAPTER 6 OF 12 · FIT AND TEST
6. Work one WLF calculation
Let Tref = 20°C, C1 = 8.0 and C2 = 100°C for a purely illustrative fit. At T = 30°C, log10 aT = −8×10/(100+10) = −0.727, so aT ≈ 0.187. Under fR = aT f, a 10 Hz measurement maps to about 1.87 Hz.
At T = 10°C, log10 aT = 80/90 = 0.889, so aT ≈ 7.74. The asymmetry arises from the denominator. These constants were invented for the example and should not be transferred to a real polymer.
CHAPTER 7 OF 12 · FIT AND TEST
7. Check predictions against held-out temperatures
Fit C1 and C2 using some temperatures, then predict shifts for a held-out curve. Compare horizontal alignment, peak position and shape residuals. A fitted equation that merely reproduces the hand-chosen shifts is not an independent validation.
Also compare WLF with an Arrhenius-type relation where physically appropriate. Model selection should follow the supported temperature regime and mechanism, not the convenience of a familiar equation.
| Check | Supports superposition | Warning sign |
|---|---|---|
| Neighbouring-curve overlap | Same shape after one horizontal shift | Systematic shoulder mismatch |
| Multiple response functions | Compatible aT values | Storage and loss require different shifts |
| Held-out temperature | Predicted shift aligns curve | Large extrapolation residual |
CHAPTER 8 OF 12 · TRANSFER AND LIMIT
8. Changed-context transfer: creep and frequency sweeps
Move from dynamic mechanical frequency sweeps to creep compliance. The stable idea is a temperature-dependent time scale. The axes, response function and direction of the shift change. A frequency convention cannot be copied blindly into a time-domain plot.
Transfer succeeds when reduced time or frequency is re-derived and overlap is checked again. It fails when one master curve is treated as proof that all relaxation modes accelerate by the same factor under every loading history.
CHAPTER 9 OF 12 · TRANSFER AND LIMIT
9. Recognise thermorheological complexity
Crystallisation, curing, physical ageing, phase separation, moisture loss or a changing morphology can alter curve shape rather than merely translate it. Different molecular mechanisms may also carry different temperature sensitivities.
Use replicate specimens and post-test characterisation where relevant. If the material itself changes during the temperature programme, the shifted curves do not represent one unchanged material observed on different clocks.
CHAPTER 10 OF 12 · TRANSFER AND LIMIT
10. Limits and common misconceptions
A master curve extends an observation window under assumptions; it does not directly demonstrate service life, fracture resistance or chemical stability. Extrapolating many decades beyond measured overlap can make a small shift-factor error enormous.
Horizontal shifting is not always sufficient. Vertical shifts may be physically justified for density or modulus effects, but extra flexibility can also hide model failure. Record every adjustment and show the unshifted data.
CHAPTER 11 OF 12 · EVIDENCE AND ROUTES
11. Questions readers often ask
The glass-transition temperature is not one immutable number. It depends on measurement method, time scale, thermal history and definition. WLF constants therefore travel with the protocol and reference state.
Does a smooth master curve prove the WLF equation?
No. Smooth overlap supports superposition over the tested range. WLF is a separate parametric description of the shift factors and needs its own residual and range checks.
Can master curves predict decades of behaviour?
They can support carefully bounded extrapolation when mechanisms remain stable. They cannot include unmeasured ageing, oxidation, damage, crystallisation or environmental change by default.
CHAPTER 12 OF 12 · EVIDENCE AND ROUTES
12. Evidence, connected routes and the next check
IUPAC’s time–temperature superposition definition states the shift-factor principle, and its WLF entry gives the equation and application range. The original Williams–Landel–Ferry 1955 paper connects amorphous-polymer relaxation to temperature.
Continue through waiting-time spectra and restart superposition, shear cessation and structural recovery, and direction-dependent entanglement recovery. The next practical check is to publish raw curves, pairwise overlaps, shift uncertainty and held-out-temperature residuals together.
Contents · Previous chapter · Return to the Polymer Science root