English is currently available.
Polymer Engineering DISCUSSION
Datasheets list a glass transition temperature for some plastics, a melting point for others and sometimes both. Polypropylene has a glass transition below 0 °C, yet it is obviously a solid at room temperature, while polystyrene softens at around 100 °C and has no melting point listed at all.
What is the physical difference between the two temperatures, and which one should I look at to decide whether a part will keep its shape at, say, 90 °C under load?
The glass transition (Tg) belongs to the amorphous, disordered part of a polymer. Below Tg the chain segments are frozen in place and the material is a rigid glass; above it the segments can move and the amorphous material becomes leathery and then rubbery. It is a gradual change in stiffness over a range of temperature, not a phase change.
Melting (Tm) belongs to crystalline regions, where chains are packed in an ordered way. It is a true first-order transition with a latent heat. Fully amorphous polymers such as polystyrene, PMMA and polycarbonate have only a Tg. Semi-crystalline polymers such as polyethylene, polypropylene, nylon and PET have both, with Tg always below Tm.
That explains polypropylene. Its Tg is around -10 to 0 °C, so at room temperature the amorphous fraction is already rubbery, but the crystals, which melt at about 165 °C, hold the structure together. The result is a tough, moderately stiff solid between Tg and Tm. Below Tg the amorphous part turns glassy as well, which is why ordinary PP becomes brittle in a freezer.
Polystyrene has no crystals to fall back on. Its Tg is about 100 °C, and once it passes that it has no structure left to carry load, so for an amorphous plastic Tg is effectively the upper limit.
Typical values, which vary with grade and test method: polystyrene Tg about 100 °C; PMMA about 105 °C; polycarbonate about 147 °C; PET Tg about 75 °C and Tm about 255 °C; nylon 66 Tg about 50 to 60 °C when dry and Tm about 260 °C; HDPE Tm about 130 °C with a Tg far below room temperature.
For a loaded part at 90 °C, amorphous candidates need a Tg well above 90 °C, which points to polycarbonate rather than polystyrene or PMMA. Semi-crystalline grades can work above their Tg, but their stiffness falls noticeably on passing it, so an unfilled nylon or PP part at 90 °C is much more flexible than at room temperature.
Neither Tg nor Tm is a design limit by itself. The practical figure on datasheets is the heat deflection temperature (HDT), the temperature at which a standard bar deflects a set amount under a stress of 0.45 or 1.8 MPa. For amorphous plastics HDT sits a little below Tg. For semi-crystalline plastics it depends heavily on reinforcement: glass fibre can lift the HDT at 1.8 MPa of a nylon from well under 100 °C to above 200 °C, close to the melting point, because the fibres and crystals carry the load after the amorphous phase has softened.
HDT is a short-term test. For long-term load at 90 °C you also need creep data at that temperature.
If you need to find the transitions of an unknown or recycled material, differential scanning calorimetry shows them directly: Tg appears as a step in the heat flow curve and melting as an endothermic peak whose area gives the degree of crystallinity. Dynamic mechanical analysis is more sensitive for Tg and also shows how much stiffness is lost across it.
Moisture and plasticisers matter. Nylon absorbs water, which lowers its Tg by tens of degrees, so a conditioned nylon part can be above its Tg at room temperature and noticeably less stiff than the dry-as-moulded value on the datasheet.
