TPU Temperature Ratings Explained: How to Use Tg, DMA, Heat Distortion, and Finished-Part Testing

Quick answer: TPU does not have one universal “service temperature.” Glass-transition temperature, DMA curves, brittle-temperature results, heat-distortion data, softening behavior, and thermal-aging results answer different questions. A useful operating window must be defined for the actual grade and part under the expected load, exposure time, motion, humidity, chemicals, and acceptance criteria.

A datasheet temperature is useful for screening. It is not a finished-part warranty. A cable jacket that flexes continuously, a static seal under compression, a roller under load, and a decorative overmold can reach different limits even when they use the same nominal TPU grade.

A generic-polymer DMA frequency–temperature sweep illustrates that measured viscoelastic response depends on temperature and test frequency. It is not TPU-specific product data.

Why one TPU temperature number is incomplete

Temperature changes several TPU behaviors at the same time:

• stiffness and flexibility;
• damping and rebound;
• creep and stress relaxation;
• impact response;
• dimensions under load;
• resistance to oxidation, hydrolysis, or chemical exposure;
• adhesion between TPU and another substrate;
• color, surface, and appearance.

The relevant limit depends on the failure mode. A visible part may fail when gloss or color changes. A seal may fail when compression recovery drops. A moving cable may fail when cold flexing initiates cracks. A loaded wheel or roller may fail through deformation long before the polymer appears to melt.

This is why a supplier’s recommended operating range should be treated as a starting hypothesis. Ask what method, duration, specimen, load, and retention criterion support that range.

Tg is a transition region, not a minimum-use temperature

The glass-transition temperature, or Tg, describes a region where molecular mobility changes substantially. Below this region, the soft segments of a polymer generally become less mobile and the material becomes stiffer. Above it, segmental motion increases.

Three cautions matter when reading Tg:

1. The reported value depends on the test. DSC, DMA, and other methods may identify different points in a broad transition.
2. The criterion matters. A DMA Tg may be reported from a storage-modulus change, a loss-modulus peak, or a tan-delta peak. These are not automatically the same temperature.
3. Application failure does not occur exactly at Tg. A TPU part may still work below a measured Tg under some loads and frequencies, or become unacceptable above it because stiffness, rebound, impact behavior, or sealing performance changed too much.

Do not convert a supplier Tg directly into a guaranteed low-temperature rating. Instead, use it to decide where to place cold-flex, impact, seal, rebound, or part-function tests.

What DMA tells a TPU buyer

Dynamic mechanical analysis applies a small oscillating deformation while temperature, frequency, or time is controlled. It commonly reports:

• Storage modulus (E′ or G′): the elastic or stiffness-related response.
• Loss modulus (E″ or G″): the energy-dissipating response.
• Tan delta: the relationship between loss and storage response, often used to discuss damping and transition behavior.

DMA is especially useful for comparing how candidate grades change across temperature. A flat-enough modulus window may be valuable where feel or support must remain consistent. A damping peak may be useful for vibration control but undesirable where high rebound is required.

Frequency is critical. A slow compression, a walking step, a high-speed roller contact, and a vibrating component do not challenge the polymer in the same way. A DMA curve measured at one frequency cannot represent every duty cycle.

Thermal-analysis instruments help characterize transitions, but they do not replace finished-part testing.

DSC, softening, melting behavior, and degradation are different events

Differential scanning calorimetry measures heat flow associated with thermal events. For TPU, the trace can be complex because grades contain hard and soft segments, may have multiple ordered domains, and retain processing and thermal history.

DSC can help compare lots, investigate transitions, or study melting and crystallization behavior. However, an endothermic feature is not automatically a service-temperature limit. The part can lose stiffness, creep, relax, or distort under load before a prominent melting event appears.

Likewise, “softening” is not the same as chemical degradation. Reversible modulus change, permanent dimensional change, oxidation, chain scission, hydrolysis, and additive migration are different phenomena. A short hot excursion may cause temporary softness; long exposure at a lower temperature may cause irreversible aging.

Illustrative DSC curve; it is not GOAFLEX TPU product data.

Heat distortion and Vicat results are not continuous-use ratings

Heat-deflection and Vicat-type tests use defined specimens, loads, heating rates, and endpoints. They are useful for standardized comparison, but they do not reproduce every TPU application.

A heat-deflection result does not answer:

• how much a seal creeps after 1,000 hours;
• whether a cable cracks after repeated hot-to-cold cycles;
• whether a wheel develops a permanent flat spot;
• whether an overmold loses adhesion to a rigid substrate;
• whether humid heat accelerates hydrolysis;
• whether oil exposure changes hardness and dimensions.

The same caution applies to short-term tensile tests at elevated temperature. A part under continuous load often needs creep or stress-relaxation testing, not only a strength value measured within minutes.

Time, load, frequency, geometry, and environment change the answer

A practical temperature review should define six variables:

Variable	Questions to specify
Temperature	Minimum, normal, maximum, peaks, ramp rate, and cycle count
Time	Seconds, hours, continuous exposure, storage life, and recovery time
Load	Tension, compression, bending, torque, internal pressure, and contact stress
Motion	Static, occasional flex, continuous flex, impact, sliding, or rolling
Geometry	Thickness, radius, notch, weld, gate, bondline, and constrained dimensions
Environment	Humidity, water, oil, fuel, cleaner, UV, oxygen, and substrate contact

Combined exposures matter. Heat can accelerate chemical aging. Water and humidity can change hydrolysis risk. Oil can plasticize or swell some constructions. A sharp radius can magnify cold-strain concentration. A rigid insert can restrain thermal expansion and create interfacial stress.

Build a cold, ambient, and hot validation matrix

A screening matrix should connect each exposure to the property that determines failure.

Exposure block	Suggested measurements	What it helps answer
Cold conditioning	Hardness/modulus, impact, bend or flex, crack inspection, part function	Does the part become too stiff or crack-prone?
Ambient baseline	Dimensions, mechanical properties, appearance, assembly force	What is the controlled reference?
Hot short-term	Dimensions under load, seal force, rebound, surface and bond integrity	Can the part tolerate peaks or processing excursions?
Hot long-term	Property retention, mass/dimensions, creep or compression set, color	Does aging or permanent deformation accumulate?
Thermal cycling	Leak/function test, crack inspection, bondline and interface checks	Do expansion mismatch and repeated transitions create damage?
Combined heat/humidity or fluid	Retained properties and finished-part function	Does the service environment accelerate degradation?

Measure both during exposure and after controlled recovery when relevant. A part may soften while hot but recover acceptably, or it may appear normal after cooling while having lost strength or adhesion.

Environmental chambers support controlled conditioning and cycling; the acceptance test must still match the part’s function.

Application examples

Cable and hose

Check cold bend or flex at the actual diameter, dynamic flex at representative frequency, hot-pressure or dimensional stability, and retained performance after oil, water, or humidity exposure. Test the complete cable or reinforced hose construction, not only a molded plaque.

Seals and gaskets

Compression set alone is not enough. Review sealing force, leakage, dimensional recovery, fluid compatibility, and thermal cycling at the actual squeeze and gland geometry.

Wheels, rollers, and damping parts

Evaluate creep, flat spotting, hysteresis heat build-up, rolling fatigue, and property retention under representative load and speed. Internal heat generation may make the part hotter than the surrounding air.

Footwear and sporting goods

Compare cold stiffness, flex-crack resistance, rebound, compression behavior, and bond durability. Test the assembled component because foam, adhesive, fabric, and geometry influence the result.

Overmolded parts

Check TPU properties together with substrate expansion, mechanical interlock, chemical adhesion, residual stress, and aging. The TPU may remain intact while the interface fails.

How to write a usable temperature requirement

Avoid specifications such as “temperature resistance: 100°C” without context. A useful requirement identifies the part, exposure, load, duration, condition during measurement, and permitted change. For example, define whether the requirement concerns a brief process excursion, continuous service, storage, thermal cycling, or performance while hot.

A robust requirement can include:

• conditioning temperature and duration;
• applied load, compression, pressure, bend radius, or motion;
• humidity or fluid exposure;
• whether the part is tested hot, after cooling, or both;
• allowed dimensional change, hardness shift, force loss, leakage, color change, cracking, or property retention;
• number of cycles and recovery period;
• complete test method, specimen, and lot-sampling plan.

This turns a broad material claim into a reproducible engineering decision. It also prevents two suppliers from quoting different “service temperatures” based on unrelated criteria.

Common interpretation mistakes

• Treating Tg as a universal minimum operating temperature.
• Treating a melting peak as the maximum safe use temperature.
• Treating heat-deflection data as a continuous-use rating.
• Ignoring test frequency when using DMA data.
• Testing unloaded specimens for a continuously loaded application.
• Reporting only post-recovery results when in-service behavior matters.
• Using one exposure temperature without cycle, ramp, or duration details.
• Assuming all TPU grades share one thermal window.
• Ignoring humidity, oil, cleaners, UV, or an attached substrate.

A practical approval workflow

1. Define the part’s thermal history and primary failure mode.
2. Shortlist grades using comparable supplier data and identical test methods.
3. Establish an ambient baseline on conditioned specimens.
4. Screen material-level behavior at cold, normal, and hot conditions.
5. Add load, motion, humidity, or chemical exposure in controlled combinations.
6. Test the real part or construction during exposure and after recovery.
7. Set numerical acceptance criteria before testing.
8. Record lot, processing history, thickness, conditioning, and test setup.
9. Repeat critical tests across representative production lots.

What to send GOAFLEX for a temperature-window review

Send the following before requesting a grade recommendation:

• application and part drawing;
• current material and hardness;
• minimum, continuous, peak, and cycling temperatures;
• exposure duration and recovery time;
• static and dynamic loads;
• motion frequency, bend radius, or compression level;
• contact with water, humidity, oil, cleaner, fuel, UV, or another polymer;
• required tests and acceptance criteria;
• current failure photos and tested samples;
• processing method, thickness, color, and substrate details.

GOAFLEX can help structure the screening and validation plan, but final approval should remain tied to the customer’s finished part, process, and service requirements. Contact hello@goaflex.com with the application details.

Image credits

• DMA frequency–temperature sweep: Cjp24, Wikimedia Commons, CC BY-SA 3.0. Converted from PNG to optimized JPEG; generic polymer data, not TPU-specific.
• Differential scanning calorimeter: U.S. government source, Wikimedia Commons, public domain.
• Illustrative DSC curve: Aboalbiss, Wikimedia Commons, CC BY-SA 3.0.
• Environmental chamber: Cjp24, Wikimedia Commons, CC BY-SA 3.0.

These search-sourced images provide testing or application context. They do not prove GOAFLEX ownership, creator endorsement, exact TPU chemistry, grade identity, or product performance.

References

• Covestro — Thermal Properties of TPU
• BASF — Elastollan Thermoplastic Polyurethane Elastomers
• TA Instruments — Measurement of Glass Transition Temperatures by DMA
• NETZSCH — Thermal Behavior of TPU by DSC
• GOAFLEX — How to Read a TPU Technical Data Sheet
• GOAFLEX — TPU Cable Jacket Selection
• GOAFLEX — TPU Seals and Gaskets