What is the difference between HDT and Vicat softening temperature?

HDT (heat deflection temperature) measures the temperature at which a specimen deflects by a specified amount under a bending load, tested according to ISO 75 or ASTM D648. The Vicat softening temperature (VST) measures the temperature at which a flat-ended needle with a 1 mm2 tip penetrates 1 mm into the surface under a fixed load, tested according to ISO 306 or ASTM D1525. HDT describes load-bearing capability; Vicat describes surface softening. Vicat is usually the higher value of the two.

Why is the Vicat softening temperature higher than the HDT?

Vicat loads only 1 mm2 of surface while the surrounding material stays rigid and constrains the loaded zone from three directions, so the polymer must soften over a wider region before the needle can sink 1 mm. In the HDT test the full cross-section carries the bending stress, so any loss of modulus anywhere produces visible deflection. The extra mechanical constraint, not a different physical property, is why Vicat reads higher.

How much higher is the Vicat temperature compared to HDT?

It depends on crystallinity. For amorphous polymers such as polycarbonate, the difference is usually about 5 to 15 degrees Celsius. For semi-crystalline polymers the gap can be 50 to 80 degrees Celsius. HDPE is an extreme example, with an HDT near 50 degrees Celsius at 1.8 MPa and a Vicat B50 near 125 degrees Celsius.

Is the HDT value the maximum service temperature of a plastic part?

No. HDT is a short-term value measured over a few minutes under constant load. Real parts experience dynamic loading, long-term thermal aging and chemical exposure, so continuous service temperature is normally 20 to 40 degrees Celsius below HDT for amorphous polymers. Use long-term indices such as the UL 746B Relative Temperature Index (RTI) together with creep data when designing for sustained high temperature.

Which ISO and ASTM standards cover HDT and Vicat testing?

HDT is covered by ISO 75-1/2/3 internationally, ASTM D648 in the United States and GB/T 1634 in China. Vicat softening temperature is covered by ISO 306, ASTM D1525 and GB/T 1633. Both tests can be run on the same HDT/Vicat tester by changing the loading head and the applied weights.

Why do two datasheets list different HDT values for the same polymer?

The applied flexural stress has to match, otherwise the numbers are not comparable. Polypropylene tested at 1.80 MPa (Method A) gives an HDT of about 57 degrees Celsius, while the same polymer tested at 0.45 MPa (Method B) gives about 99 degrees Celsius. Moisture content, annealing state and specimen orientation also shift the result, and most datasheets report only dry-as-molded values.

When should I use HDT instead of the Vicat softening temperature?

Use HDT when the part carries sustained mechanical load at elevated temperature, for example automotive intake manifolds, structural brackets, electrical enclosures and connectors. Use Vicat when the concern is dimensional stability without significant load, such as processing and demolding windows, thin-wall or film performance, incoming material inspection, and hot-water stability of PVC or CPVC pipe.

How does glass fiber reinforcement change the HDT?

The effect depends on whether the polymer is semi-crystalline. PA66 goes from an HDT of about 85 degrees Celsius unfilled to about 250 degrees Celsius with 30 percent glass fiber, approaching its melting point of 260 degrees Celsius. Polycarbonate, which is amorphous, only moves from about 135 to 145 or 150 degrees Celsius with the same fiber loading, because there is no crystalline phase left to help the fibers carry the bending load above Tg.

How do HDT and Vicat relate to glass transition temperature (Tg) and melting point (Tm)?

In amorphous polymers there is only one thermal transition, so HDT and Vicat both sit close to Tg and within about 10 degrees Celsius of each other. In semi-crystalline polymers the crystalline regions keep supporting stiffness above Tg, which creates a plateau, so the values follow the order Tg below HDT, HDT below Vicat, and Vicat below Tm.

Which Vicat method should be used when comparing materials?

Use method B50, meaning 50 N at 50 degrees Celsius per hour, which is the default specified by ISO 10350 for comparable plastics data. A higher load gives a lower result, and the faster 120 degrees Celsius per hour rate can read up to 10 degrees higher than the 50 degrees Celsius per hour rate because the polymer has less time to relax. Comparing results across different methods leads to false conclusions.

Does molded-in residual stress affect HDT and Vicat results?

Yes for HDT and much less for Vicat. Three-point bending involves the whole cross-section, so frozen-in stress from injection molding relaxes during heating and adds deflection that is unrelated to heat resistance, lowering the measured HDT. The Vicat needle affects only a tiny local volume that is constrained by surrounding cold material, so VST is relatively insensitive to processing history, which makes it more repeatable for quality control and incoming inspection.

Why does moisture reduce the HDT of nylon?

Absorbed water plasticizes the amorphous phase of polyamide and lowers its glass transition temperature, so stiffness drops at a lower temperature. Dry-as-molded PA66 has an HDT of roughly 85 degrees Celsius at 1.8 MPa, while a part conditioned to equilibrium humidity can drop to around 60 degrees Celsius. Datasheet values are usually measured dry, so designs that cannot guarantee a dry part should be sized using conditioned-state data.

HDT vs Vicat Softening Temperature: How to Read Both Numbers

Open almost any thermoplastics datasheet and you will find two numbers sitting next to each other under “thermal properties.” One is heat deflection temperature, HDT for short. The other is Vicat softening temperature, usually abbreviated VST. They are printed in adjacent cells, both in degrees Celsius, and people read them as two flavors of the same thing.

They are not the same thing, and treating them as synonyms is how parts end up sagging in service.

HDT asks: how hot can this loaded part get before it bends too far? Vicat asks: how hot can this surface get before something starts sinking into it? One is about structure holding load. The other is about a surface losing its nerve. Pick the wrong one for your decision and you can be off by 75 °C, which is not a rounding error.

This guide covers what each test does, why Vicat nearly always reads higher, how crystallinity decides the size of the gap, why glass fiber transforms one family of polymers and barely touches the other, and which number belongs in your specification.

The 60-second version

Both are short-term heat tests. They load the polymer in completely different ways.

HDT bends a bar. A specimen rests on two supports, gets a steady flexural load at the midpoint, and sits in a silicone oil bath warming at 120 °C/h until it deflects by a set amount. It simulates a structural part doing structural work. ISO 75 and ASTM D648 are the governing standards.

Vicat stabs a surface. A flat-tipped hardened steel needle with a 1 mm² contact area presses down under a fixed weight while the bath heats at 50 or 120 °C/h. When the needle has sunk 1 mm, you write down the temperature. It simulates surface softening. ISO 306 and ASTM D1525 cover it.

Because Vicat measures local penetration and HDT measures the deflection of an entire beam, Vicat comes out higher. For amorphous materials like polycarbonate the difference is modest, roughly 10 to 15 °C. For semi-crystalline ones it can be staggering. HDPE shows an HDT around 50 °C at 1.8 MPa and a Vicat near 125 °C.

Neither figure is a material property in the way density is. Both depend on load, heating rate, specimen shape, and how the part was molded. Neither one is your maximum service temperature.

HDT: what happens when you heat a loaded bar

Start with the definition. HDT is the temperature at which a standard bar deflects by a specified amount under a specified bending stress. A cleaner way to picture it: it is one point on the modulus-versus-temperature curve, the point where stiffness has dropped enough that the load wins.

The setup

The bar goes on two supports, gets loaded in the middle, and soaks in a silicone oil bath that climbs at 120 °C/h. Using a liquid bath rather than hot air removes variability in heat transfer, which is why the number is reproducible across labs.

Endpoints differ by standard:

ISO 75 measures a deflection equal to a 0.20 % increase in outer fiber strain.
ASTM D648 uses a flat 0.25 mm of deflection.

The load is set by choosing a stress level:

Method	Flexural stress	Intended use	Typical materials
A	1.80 MPa (ISO) / 1.82 MPa (ASTM)	High stress, rigid plastics	PC, PEEK, PA, PPS
B	0.45 MPa (ISO) / 0.455 MPa (ASTM)	Low stress, flexible plastics	PP, PE, PTFE
C	8.00 MPa (ISO only)	Very high stress, hard compounds	Hard rubber, heavily filled systems

Datasheets usually quote Method A. Usually. Check anyway, because plenty of them do not say.

The load trap, which ruins more comparisons than anything else

Same polypropylene, two methods:

Method A, 1.80 MPa: HDT about 57 °C
Method B, 0.45 MPa: HDT about 99 °C

Forty-two degrees of separation from one polymer. Now imagine building a material shortlist by pulling whichever value each supplier printed. Your ranking would be fiction. There is no other single habit that damages material selection quite this efficiently.

ISO 75 against ASTM D648

The headline difference is orientation. ISO lays the specimen flat. ASTM stands it on edge.

Detail	ISO 75	ASTM D648
Specimen	80 × 10 × 4 mm	127 × 12.7 × 6.4 mm
Orientation	Flatwise	Edgewise
Span	64 ± 1 mm	101.6 mm (A) / 100.0 mm (B)
Heating rate	120 ± 10 °C/h	2 ± 0.2 °C/min
Start temperature	Below 27 °C	Ambient
Endpoint	0.20 % strain increase	0.25 mm

Edgewise changes the second moment of inertia, so the hanging weight differs to reach the same fiber stress. In practice the two standards land within 2 to 5 °C of each other, which most of us treat as equivalent. The variables that genuinely move the result are the load level, how much water the nylon has absorbed, and whether the part was annealed.

What HDT is not

It is not the temperature where softening begins. It is the temperature where deflection hits a target value under a target load. Those sentences describe different physical events. A polymer can be past its Tg and still post a healthy HDT if crystals or fibers are picking up the load.

It is also not creep data. The whole test is over in minutes. Anything expected to hold a continuous load at temperature for months needs creep and creep-rupture curves, and HDT will not substitute for them.

Vicat: one square millimeter under a needle

The Vicat test borders on crude, and that is the point. A hardened steel needle with a flat circular tip of 1 mm² sits on the specimen. A weight pushes down. The assembly warms in an oil bath. When penetration reaches 1 mm, record the temperature. Three moving parts, no bending rig, no strain calculation.

Four ways to run it

ISO 306 defines methods by combining load and heating rate:

Method	Load (N)	Rate (°C/h)	When you use it
A50	10	50	Slow and gentle. Best repeatability.
B50	50	50	The default under ISO 10350
A120	10	120	Quick screening at low load
B120	50	120	Quick screening at high load

Two behaviors follow directly from the mechanics.

Heavier load means the needle gets there sooner, so B methods read lower than A methods on identical material. Faster heating means less time for the polymer to relax under the tip, so 120 °C/h results can come in up to 10 °C above their 50 °C/h counterparts.

Cross-supplier comparisons only make sense when the method matches. ISO 10350, the standard for collecting comparable plastics data, settles on B50: 50 N at 50 °C/h.

Requirements worth knowing

Thickness between 3 and 6.5 mm, minimum 10 × 10 mm of surface.
Molded bars or pieces cut from sheet and finished parts both work.
Thin sheet can be stacked, three layers maximum, at least 3 mm total.
Needle diameter 1.128 ± 0.008 mm, with vertical travel accurate to 0.02 mm.
Minimum immersion depth 35 mm. Start at least 50 °C under the expected result.
Temperature sensor within 12.5 mm of the needle tip.
Two specimens minimum per condition, and rerun if they disagree by more than 2 °C.

The part where Vicat quietly beats HDT

HDT involves the full cross-section, so frozen-in molding stress gets involved too. Heat the bar, let that internal stress relax, and it contributes deflection that has nothing to do with heat resistance. The number drifts with processing history.

The needle touches almost nothing. Surrounding material stays rigid and clamps the loaded zone from three sides, so residual stress has little room to act. VST tracks the polymer itself rather than the mold. That is why it dominates incoming inspection even in shops where HDT governs the design, and why suppliers like it for certificates of analysis.

Side by side

Aspect	HDT	Vicat (VST)
Loading	Three-point bending, distributed	Point load through a needle
Load	0.45 to 1.80 MPa flexural	10 to 50 N concentrated
Endpoint	0.25 mm or 0.20 % strain	1 mm penetration
Detects	Loss of stiffness in bulk	Surface softening
Residual stress	Affects the result	Barely matters
Relative size	Lower	5 to 15 °C higher typically, far wider for semi-crystalline grades
Standards	ISO 75, ASTM D648, GB/T 1634	ISO 306, ASTM D1525, GB/T 1633
Use it for	Loaded structural parts	Processing windows, surface stability, QC

So why is Vicat the bigger number?

The answer is constraint.

In bending, every fiber across the section carries load. Warm the bar and modulus falls everywhere at once. Softening in any layer drags the whole beam down, and there is nowhere to hide. Deflection shows up early.

Under the needle, only 1 mm² is being pushed, and everything around it is still cool and stiff. That surrounding material squeezes from three directions. To sink a full millimeter, the needle needs a decent volume of polymer to give way, not just a hot spot. More heat required, higher number recorded.

No mystery, no contradiction between the two tests. Just different amounts of material being asked to move. The effect magnifies whenever the polymer has internal structure resisting progressive heating, which takes us to crystallinity.

The trap worth committing to memory

With HDPE, PP, PBT and their relatives, gaps of 50 to 80 °C between the two values are routine rather than exceptional. HDPE sits near 50 °C on HDT at 1.8 MPa and near 125 °C on Vicat B50. Using one in place of the other is not a conservative shortcut. It is simply wrong, and the error is large enough to fail a product that looked perfectly specified on paper.

Crystallinity runs the whole show

Everything about the relationship between HDT, Vicat, Tg and Tm comes down to one question: does your polymer have crystals?

Amorphous: everything collapses at once

PC, PMMA, PSU, PEI, PAI and friends have a single thermal event. Cross Tg and the material goes from glassy to rubbery, with modulus falling three to four orders of magnitude. Both tests trigger near that drop because it is the only drop available.

Patterns to expect:

HDT lands 5 to 15 °C under Tg, since deflection begins the moment modulus starts sliding.
Vicat lands at or a little above Tg.
The two values sit within about 10 °C of each other.
Glass fiber does very little. Once the matrix goes rubbery, there is no crystalline framework left to help the fibers carry a bending load.

Polycarbonate spells it out: Tg near 145 °C, HDT about 135 °C, Vicat about 150 °C.

Semi-crystalline: a plateau held up by crystals

PA, POM, PP, PBT, PEEK, PPS and PET have two transitions. Amorphous regions soften at Tg, but the crystals stay intact and keep working as scaffolding. Modulus drops and then flattens. Somewhere across that plateau the remaining stiffness finally falls under the test threshold.

The progression:

Tg arrives, amorphous regions soften, modulus declines but remains real
Plateau where HDT and Vicat live, with crystals carrying the load
Tm arrives, crystals melt, stiffness disappears

What that produces:

HDT somewhere between Tg and Tm, exact position set by how much crystallinity there is.
Vicat typically 20 to 80 °C above HDT, since a needle needs some crystal softening too.
Enormous glass fiber response.
Higher crystallinity pushes both numbers toward Tm.

PEEK is the clean example. Tg of 143 °C, Tm of 341 °C. Unfilled, it manages an HDT near 160 °C, barely above Tg, thanks to something like 30 to 35 % crystallinity. Add 30 % glass fiber and HDT reaches about 315 °C. Almost to the melting point, from a resin whose Tg is 143 °C.

If you remember one line, make it this one. Amorphous: HDT ≈ Vicat ≈ Tg. Semi-crystalline: Tg < HDT < Vicat < Tm. That explains why PC’s pair nearly overlap while PP’s sit 30 to 50 °C apart.

Glass fiber: spectacular in one case, nearly useless in the other

This asymmetry deserves more attention than it gets, and it has cost people real money.

Semi-crystalline grades make a jump toward Tm

PA66 has a dry Tg around 50 °C and melts near 260 °C. Unfilled, its HDT is roughly 85 °C. Add 30 % glass fiber and HDT reaches about 250 °C. That is a 165 °C gain, moving the material from “slightly better than Tg” to “almost at Tm.”

PA6 tells the same story, with published figures at 1.82 MPa:

Unfilled: HDT 65 °C, modulus 2.81 GPa
14 % glass fiber: HDT 200 °C, modulus 4.46 GPa
33 % glass fiber: HDT 210 °C, modulus 7.87 GPa

Note what did not move. Tg goes from about 65 to 70 °C. The crystals do not change. What changes is load transfer: past Tg the softened matrix hands the work to the fibers, which stay rigid all the way up to melting. In PA, PP and PBT, 10 wt % of short glass fiber is often enough to push HDT close to Tm.

Amorphous grades shrug

Unfilled PC: HDT about 135 °C at 1.8 MPa
30 % glass-filled PC: HDT about 145 to 150 °C

Ten or fifteen degrees for a third of the part weight in fiber. The reason traces back to the missing plateau. Above Tg there is nothing to keep the fibers working as a load path, so the section deflects as a unit. Stiff fibers floating in a rubbery matrix cannot prevent whole-beam bending.

If higher heat resistance is what you need from PC, switching to a higher-Tg resin gets you there. Loading it with fiber mostly gets you a higher price per kilo.

Aspect ratio sets the ceiling

The rule extends past glass fiber to every filler:

Filler	Aspect ratio	Effect on HDT
Glass bead, calcium carbonate	~1, spherical	Minimal
Talc, kaolin	10 to 40:1	Moderate
Wollastonite, wet-ground mica	10 to 50:1	Good
Chopped glass fiber, nanoclay	50 to 200:1	Strong
Carbon fiber, aramid fiber	100 to 500:1	Strongest

Screening data across common engineering plastics

Use this for shortlisting. Do not use it for design.

Material	Type	Tg (°C)	Tm (°C)	HDT @ 1.8 MPa (°C)	Vicat B50 (°C)
HDPE	Semi-crystalline	−110	135	~50	~125
PP	Semi-crystalline	−10	165	~57	~90
PP + 30% GF	Reinforced semi-crystalline	−10	165	~150	~160
Rigid PVC	Amorphous	80	n/a	~70	~85
PS (GPPS)	Amorphous	100	n/a	~85	~95
PMMA	Amorphous	105	n/a	~95	~110
ABS	Amorphous	105	n/a	~95	~100
PET	Semi-crystalline	75	255	~80	~170
PET + 30% GF	Reinforced semi-crystalline	75	255	~230	~250
PBT	Semi-crystalline	45	225	~60	~165
PBT + 30% GF	Reinforced semi-crystalline	45	225	~210	~220
POM (copolymer)	Semi-crystalline	−60	175	~110	~150
PA66 (dry)	Semi-crystalline	50	260	~85	~240
PA66 + 30% GF	Reinforced semi-crystalline	50	260	~250	~255
PC	Amorphous	145	n/a	~135	~150
PC + 30% GF	Reinforced amorphous	145	n/a	~148	~155
PSU	Amorphous	187	n/a	~175	~188
PEI	Amorphous	217	n/a	~200	~215
PPSU	Amorphous	220	n/a	~205	~218
PVDF	Semi-crystalline	−35	171	~112	~155
PTFE	Semi-crystalline	−97	327	~55	~110
PPS	Semi-crystalline	90	280	~135	~200
PPS + 40% GF	Reinforced semi-crystalline	90	280	~260	~270
PEEK	Semi-crystalline	143	341	~160	~250
PEEK + 30% GF	Reinforced semi-crystalline	143	341	~315	~330
PEEK + 30% CF	Reinforced semi-crystalline	143	341	~315	~340
PAI	Amorphous	275	n/a	~275	n/a
PI	Amorphous	360	n/a	~360	n/a

Three things this table gives away

The semi-crystalline rows have wide gaps. PA66 reads 85 °C against 240 °C. PC reads 135 against 150.

Glass fiber gains are wildly uneven between families. PA66 gains 165 °C. PC gains about 13.

PTFE needs a warning label. Continuous-use temperature of 260 °C, HDT of around 55 °C. It loves heat and hates load. Nothing about those two facts contradicts, but put PTFE into a hot load-bearing job on the strength of its reputation and it will creep into failure.

Which one does your application need

HDT, when something is carrying load

Go to HDT when:

The part holds sustained mechanical load at temperature, whether bending, compression or vibration.
You want a short-term ceiling for structural service temperature.
The application is under-hood automotive, an electrical enclosure, a loaded bracket, or comparable.
You are verifying batch-to-batch structural heat performance in QC.

Typical cases: glass-filled PA intake manifolds, PC/ABS appliance housings, PBT and LCP connectors, PEEK and PPS structural components.

Vicat, when the surface is the concern

Vicat fits when:

Heat arrives without meaningful load, but shape has to hold.
You are setting a demolding window. It belongs comfortably below VST.
You are running incoming inspection, where immunity to residual stress pays off.
Thin walls, films and coatings are what you are judging.

Typical cases: hot-water dimensional stability in PVC and CPVC pipe, where specifications such as IS 4984 call for VST of at least 125 °C; interior trim baking in a parked car; hot-fill packaging; cable insulation.

Four cautions

HDT is not your service ceiling. Real parts face dynamic loading, years of thermal aging, and whatever chemicals are in the neighborhood. For amorphous polymers especially, take 20 to 40 °C off the HDT as a starting figure for continuous use, then confirm against long-term data such as the UL 746B Relative Temperature Index.

Annealing helps. Fast cooling during molding locks in stress and holds crystallinity down, and both suppress measured HDT. Heat above Tg but below Tm, cool slowly, and you relieve the stress while growing crystals. HDT rises, dimensional stability improves.

Nylon drinks water. Absorbed moisture plasticizes the amorphous phase and pulls Tg down with it. Dry-as-molded PA66 sits around 85 °C. Conditioned to equilibrium humidity it can fall to roughly 60 °C. Published values are almost always dry, so unless you can guarantee the part stays dry, design for the conditioned state.

Stop substituting one for the other. Semi-crystalline grades can differ by 50 to 80 °C between the two tests. Ask for both on every datasheet you receive.

Standards reference

Property	International	US	China
HDT	ISO 75-1/2/3	ASTM D648	GB/T 1634
Vicat	ISO 306	ASTM D1525	GB/T 1633

One machine covers both tests. Swap the loading head, change the weights, pick your method, and start the bath. Sharing the frame and the heating system is why so many labs quote them together. Current testers carry two to six stations, so a batch runs in a single cycle.

Where this leaves you

Two tests, two failure modes, two questions. HDT cares whether a loaded part still resists deflection. Vicat cares whether a hot surface still resists penetration. Neither tells you a safe long-term operating temperature by itself, and neither can stand in for the other.

Work backward from the part. Loaded at temperature? HDT at the right stress level is your figure, supported by creep data if the loading never stops. Worried about surface deformation, demolding, or batch verification? Vicat, with its method stated alongside. Then check amorphous versus semi-crystalline, because that decides how wide the gap should be and whether throwing glass fiber at the problem will accomplish anything.

Reading these numbers off a datasheet takes about a minute. Knowing which one applies takes slightly longer, and it is the part that actually matters.

Next steps worth taking:

Shortlist on UL 746B RTI values rather than HDT when the design has to survive years of heat.
Ask every supplier for Method A and Method B HDT, plus the Vicat method, before comparing quotes.
For nylon and other hygroscopic grades, request conditioned-state properties, not just dry-as-molded.