Use heat deflection temperature, not glass transition temperature alone, when selecting engineering plastics for loaded service: ASTM D648 commonly applies 0.455 or 1.80 MPa while recording a specified 0.25 mm deflection [S2][S5].
The comparison covers amorphous and semicrystalline thermoplastics, glass-fiber reinforcement, polymer processing history, and the limits of using a laboratory HDT value as a finished-part rating.
Tg and HDT measure different thermal limits
Glass transition temperature, Tg, marks the temperature region where molecular mobility produces a significant change in the mechanical behavior of a polymer’s amorphous phase; below it, the material is relatively rigid and glassy, while above it the material becomes softer and more compliant [S2]. Tg is not a melting point, and for semicrystalline polymers it is distinct from the crystalline melting temperature.
Heat deflection temperature is the temperature at which a standardized specimen reaches a specified deformation under a defined flexural load during controlled heating [S2]. The test measures the ability of a defined geometry to retain stiffness under load, not the temperature at which the polymer chemically changes or flows.
The relationship is not fixed. For amorphous polymers, HDT is commonly below Tg [S6].
Test method, stress, and geometry control the HDT result
A reported HDT is not a grade-independent material constant.
A material tested at 0.455 MPa will generally deflect at a lower temperature than the same material tested at 1.80 MPa because the lighter load permits more compliance.
The temperature at which the standard test specimen deflects by 0.25 mm is recorded [S5]. The 0.25 mm criterion is a test convention, not a universal allowable service deformation [S5]. Thick parts also develop a temperature gradient through the section, so a small-specimen result can overstate the performance of a large component [S5].
Material selection by load, geometry, and temperature
For short-term loaded service, compare materials at the same HDT method, stress, specimen orientation, and conditioning state.
For rigid, dimensionally stable components, select a material whose HDT exceeds the maximum service temperature with a defined engineering margin, then verify that the actual stress does not exceed the test stress. Polycarbonate and ABS are reasonable starting points for moderate heat exposure, while PEEK and PEI fit higher-load applications; reinforced grades can change the result substantially.
Do not select by Tg alone. Tg is valuable for screening rigidity loss and for understanding processing or distortion behavior, but it does not quantify deformation under a specified engineering load. For a part with a tight tolerance, spring force, or long-term clamp, use creep data or a temperature-dependent stress-strain model in addition to HDT.
Reinforcement and processing history can shift the result
Glass-fiber reinforcement can raise stiffness and heat-deflection resistance because the fibers carry load and restrict deformation. For reinforced materials, check glass fiber content, orientation, and resin grade because fiber direction and interfacial quality determine whether the test bar transfers that laboratory performance to the part.
Processing history also changes the result. Molecular weight, additives, moisture, residual species, degree of curing, and crosslink density can alter Tg and thermomechanical behavior [S2]. For 3D-printed photopolymers, wavelength, light power, exposure time, effective energy dosage, and post-curing conditions affect the network and the measured properties [S2]. An engineering photopolymer selection can therefore depend on the complete manufacturing specification, not just a resin datasheet value.
For amorphous polymers, HDT normally falls below Tg because stiffness falls through the glass-transition region before the severe deformation threshold is reached [S6]. Semicrystalline polymers can behave differently because crystalline domains constrain flow.
Thermosets, semicrystalline plastics, and photopolymers need separate treatment
Thermoset photopolymers do not normally melt like crystalline thermoplastics; they soften or lose stiffness according to their network structure, crosslink density, Tg, and processing history [S2]. Their HDT therefore depends heavily on conversion achieved during printing and post-curing, and a low-viscosity resin alone does not establish the final part’s thermal rating.
Semicrystalline grades must also be assessed above Tg and below the melting temperature. Standard PLA is reported with Tg of 55–60°C and HDT of 50–55°C [S4]. Annealing at 80–110°C can increase reported crystallinity from 10–30% to 40–50% and HDT from 50–55°C to 90–110°C for the processed PLA described, but shrinkage accompanies crystallization and must be included in the tolerance and distortion review [S4].
Use Tg to describe the transition into a softer, more mobile state and HDT to screen loaded thermal performance. Use long-term creep, chemical compatibility, fatigue, and the actual part geometry before releasing a component for sustained service.
Specify the material with a complete thermal test record
A purchase or drawing specification should state polymer family, manufacturer, grade, color or additive package, reinforcement percentage, HDT test standard and edition, flexural stress, specimen configuration, deflection criterion, conditioning, and orientation. The values in the cited comparison table are representative published figures, not universal properties: PEEK is shown at 260°C HDT under 0.455 MPa and 143°C Tg, while PEI is shown at 216°C and 215°C [S3].
For final parts, require a representative geometry test when wall thickness, fiber orientation, weld lines, print orientation, or post-processing can shift stiffness. The engineering plastic grade must therefore be approved against the complete process, not merely against a nominal resin family. Keep the test report, material certificate, conditioning history, and finished-part validation under the same revision.
The most reliable next check is to obtain HDT at the maximum design stress, then validate the production geometry at the maximum service temperature and required duration; if the design depends on sustained loading, add creep data before approval.