Contact Mechanics · Friction heating
Flash temperature: how hot a sliding contact gets
Start with the first picture: raise the sliding speed from 0.01 to 10 m/s and watch the hot spot under the slider stretch into a tail, and the flash temperature climb by more than a hundred times.
How hot does a sliding contact get?
Friction turns work into heat right at the contact. The heat per second is Q = μ W V (friction coefficient × load × speed), and it comes out on a small area of radius a. The flash temperature is how far that small area heats above the bulk. The heat is shared between the slider, which always sees the hot spot in the same place, and the flat, which sees it pass by. The Peclet number L = V a / 2κ (κ = thermal diffusivity, how fast heat spreads) says whether the heat spreads out ahead of the moving spot (L below 0.1) or is left behind in a thin tail (L above 5).
Try it: with a 6 mm steel ball on steel at 10 N and μ = 0.6, move the speed from 0.01 m/s to 1 m/s: the flash temperature goes from 2.8 °C to 185 °C, and the heat going into the flat rises from 50% to 67%. Then choose PTFE as the slider: almost all the heat goes into the steel flat.
Why are asperity flashes so much hotter than the bulk?
Real surfaces touch only at the tips of their roughness, the asperities. Each tip carries a load at a pressure close to the hardness, so a small patch takes a large share of the heat. A patch lives only while two tips pass each other, t = 2r/V, often a few microseconds. Meanwhile the whole contact, the bulk (or nominal) temperature, warms slowly as heat flows back into the pin and the disc.
Try it: at the defaults (steel on steel, 10 N, 1 m/s, 3 mm pin, r = 5 µm) each flash lasts 10 µs and adds 167 °C, while the bulk has risen less than 1 °C after 10 ms and reaches 23 °C only after about 20 s. Raise r to 10 µm: fewer, larger contacts (5 instead of 18), and each flash gets hotter (317 °C).
What does the flash temperature do to the surface?
A few hundred degrees for a few microseconds is enough to change a surface. Steel grows oxide much faster (the start of oxidative wear; choose mode 4 there), a hardened steel can lose its temper, an oil film can break down and the surfaces weld (scuffing), and PTFE can melt and smear. The ladder compares your contact temperature with these limits for the two materials chosen in the first picture.
Try it: with steel on steel (picture 1 defaults) raise the speed to 2 m/s: the contact passes the 180 °C tempering limit of hardened bearing steel. At 10 m/s it passes 727 °C, where steel turns to austenite and can leave a hard white layer. Now choose PTFE as the flat in picture 1, set μ to 0.2 and the speed to 30 m/s: the contact reaches about 430 °C, past the 327 °C melting point of PTFE.
What to take away
Heat is μWV
The heat grows with load, speed and friction. Halve the friction with a lubricant and the flash temperature halves too.
Speed changes the shape
At low Peclet number heat spreads in all directions. At high Peclet number the flat carries it away in a thin tail, and the rise grows only as √V.
Flashes are hot and short
Asperity contacts reach hundreds of degrees for microseconds. A thermocouple in the pin reads only the bulk.
Poor conductors run hot
At 10 N and 1 m/s, titanium on titanium reaches 622 °C where steel on steel reaches 185 °C and aluminium on aluminium 46 °C.
More detail: the equations and the limits of these models
Archard's two limits
Blok (1937) and Jaeger (1942) solved for a heat source moving over a half-space. Archard (1959) put the results in a simple form for a circular contact of radius a with uniform heat Q. If all the heat goes into a body that sees the source standing still (the slider), the mean rise is θ1 = Q / 4ak1. For the body the source moves over (the flat), with L = Va/2κ2: at L < 0.1 the same Q / 4ak2, and at L > 5 θ2 = 0.31 (Q / ak2)(κ2/Va)1/2. Between them the page uses θ2 = (Q/ak2) / √(16 + 20.8 L), which meets both limits and stays within 7% of each at L = 0.1 and L = 5.
Heat partition
Blok's rule: the two surfaces must have the same temperature where they touch. If a fraction f goes into the slider, f θ1 = (1 − f) θ2, so f = θ2/(θ1 + θ2) and the flash temperature is θ = θ1θ2/(θ1 + θ2). The colour picture is the exact field of a uniform disc source (the moving point source of Rosenthal and Jaeger, added up over the disc), scaled so its mean over the contact equals the Archard value. Without scaling, the exact disc mean differs from Archard's formulas by at most about 8% (at L = 0, 0.270 against 0.25 Q/ak). The peak sits at the trailing edge and is 1.2 (low L) to 1.65 (high L) times the mean. Both bodies are treated as half-spaces; the slider's field is drawn inside the ball or pin.
Contact size
For a ball, Hertz gives a = (3WR/4E*)1/3. If the mean pressure would exceed the hardness H of the softer body, the page uses the fully plastic size a = (W/πH)1/2 instead. Asperity contacts in picture 2 are always taken as plastic, at the hardness of the softer body, so their number is N = W/(πr2H).
Bulk temperature
The bulk model follows Lim and Ashby (1987): heat Q leaves the apparent contact of area An by conduction along an equivalent length l into both bodies to a heat sink at the ambient temperature, so the final rise is Q l / An(k1 + k2). The rise with time is the solution for a rod heated at one end and held cold at the other (Carslaw and Jaeger), with a diffusivity averaged over the two bodies. It sets the time scale, about l2/κ, not the exact curve. In a real tester the bulk temperature depends on the machine, the disc track being reheated every turn, and air cooling; measure it if you can.
A flash at one spot
The spike shape in picture 2 is the one-dimensional solution for a heat flux switched on for a time tc: the rise grows as √t while the contact lasts, then falls as √t − √(t − tc). The height is the Archard flash temperature of one asperity contact. The flashes are placed at random times with the right average rate: a spot is in contact for a fraction Ar/An of the time.
Limits
The thermal constants are room-temperature values and are held fixed. In alumina the conductivity falls to about half by 400 °C (Incropera), so hot ceramic contacts run hotter than shown. The heat is taken as uniform over the contact; a Hertz pressure profile changes the numbers by about 10 to 20%. μ is an input here, but it often changes with temperature (see why friction is not a material constant). No surface can get much hotter than the melting point of the lower-melting body: the molten film lowers the friction (Bowden and Tabor).
| Material | k W/m K | ρ kg/m³ | c J/kg K | κ mm²/s |
|---|
On this site: Asperities and the real area of contact · Wear mechanisms and wear depth (mode 4: tribo-oxidative wear, driven by the flash temperature) · Why friction is not a material constant · Lubrication regimes and the Stribeck curve · Wear rate calculator · Nanoindentation hardness (the H that sets asperity contact size) · Scratch testing: critical loads · Strengthening mechanisms (why tempering softens steel)
Questions people ask
What is flash temperature in tribology?
The short, local temperature rise at a sliding contact above the temperature of the bodies around it. It comes from friction heat released on a very small area. At asperity contacts it can reach several hundred degrees for a few microseconds.
What is the difference between flash temperature and bulk temperature?
The bulk (or nominal) temperature is the average temperature of the contact region and the parts near it. It rises slowly, over seconds to minutes. The flash temperature is the extra rise at each real contact spot, on top of the bulk, lasting only as long as the spot.
What is the Peclet number in sliding contact?
L = Va/2κ: sliding speed times contact radius, over twice the thermal diffusivity. It compares how fast the contact moves with how fast heat spreads. Below about 0.1 the heat spreads as if the source stood still; above about 5 it stays in a thin layer behind the moving contact.
How do you measure flash temperature?
It is hard, because the spots are small and short-lived. Bowden and Tabor used the two metals as a thermocouple. Others use infrared cameras looking through a sapphire or diamond disc, and Raman or thin-film sensors. A thermocouple in the pin measures only the bulk.
How is frictional heat divided between two sliding bodies?
Blok's rule: the share is set so both surfaces reach the same temperature at the contact. The body that conducts better, and the one that moves past the contact (it keeps meeting cold material), takes more of the heat.
What causes scuffing?
Scuffing is sudden welding and tearing of lubricated surfaces, often in gears and cams. Blok proposed that it starts when the contact temperature (bulk plus flash) reaches a critical value for the oil and its additives, where the protective films stop working.
Why does PTFE have a PV limit?
The heat μWV grows with pressure times speed (PV). PTFE conducts heat about 190 times worse than steel, so its surface heats quickly. Near its 327 °C melting point the surface softens and smears, and the wear rate rises sharply.
What temperature causes oxidative wear of steel?
Oxide grows at any temperature, but the growth rate rises steeply with temperature. Quinn's model uses the contact temperature to pick the oxide: α-Fe₂O₃ below about 450 °C, Fe₃O₄ from 450 to 600 °C and FeO above 600 °C. Flash temperatures of a few hundred degrees are enough.
References
Show the 13 references
- H. Blok, Theoretical study of temperature rise at surfaces of actual contact under oiliness lubricating conditions, Proceedings of the General Discussion on Lubrication and Lubricants, Institution of Mechanical Engineers, London, vol. 2, 222 to 235 (1937).
- J. C. Jaeger, Moving sources of heat and the temperature at sliding contacts, Journal and Proceedings of the Royal Society of New South Wales 76, 203 to 224 (1942).
- J. F. Archard, The temperature of rubbing surfaces, Wear 2, 438 to 455 (1959). doi:10.1016/0043-1648(59)90159-0
- M. Kalin, Influence of flash temperatures on the tribological behaviour in low-speed sliding: a review, Materials Science and Engineering A 374, 390 to 397 (2004). doi:10.1016/j.msea.2004.03.031
- I. M. Hutchings and P. Shipway, Tribology: Friction and Wear of Engineering Materials, 2nd ed., Butterworth-Heinemann (2017): section 3.6, frictional heating and flash temperature; chapter 5, oxidative wear.
- F. P. Bowden and D. Tabor, The Friction and Lubrication of Solids, Part I, Clarendon Press, Oxford (1950): chapter 2, the surface temperature of rubbing solids.
- H. S. Carslaw and J. C. Jaeger, Conduction of Heat in Solids, 2nd ed., Oxford University Press (1959): the disc source and the heated rod.
- S. C. Lim and M. F. Ashby, Wear-mechanism maps, Acta Metallurgica 35, 1 to 24 (1987). doi:10.1016/0001-6160(87)90209-4
- T. F. J. Quinn, Review of oxidational wear, Tribology International 16, 257 to 271 and 305 to 315 (1983).
- X. Tian and F. E. Kennedy, Maximum and average flash temperatures in sliding contacts, Journal of Tribology 116, 167 to 174 (1994). doi:10.1115/1.2927035
- T. L. Bergman, A. S. Lavine, F. P. Incropera and D. P. DeWitt, Fundamentals of Heat and Mass Transfer, 7th ed., Wiley (2011): Table A.2, polycrystalline aluminium oxide (k = 36.0 W/m K, ρ = 3970 kg/m³, c = 765 J/kg K at 300 K).
- Chemours, Teflon PTFE fluoropolymer resin: properties handbook: PTFE k = 0.25 W/m K, specific gravity 2.16, specific heat 1.2 kJ/kg K at 40 °C, transitions at 19 °C and 30 °C, melting (gel) point 327 °C, continuous service 260 °C.
- Metal data sheets: AISI 52100 (AZoM, k = 46.6 W/m K, ρ = 7810 kg/m³, melting 1424 °C; c = 475 J/kg K from MatWeb); C93200 bearing bronze (Wieland Concast data sheet: k = 58.2 W/m K, ρ = 8910 kg/m³, c = 377 J/kg K, solidus 854 °C, E = 100 GPa, 65 HB); 6061-T6 (ASM Handbook vol. 2, 1990: k = 167 W/m K, ρ = 2700 kg/m³, c = 896 J/kg K, solidus 582 °C); Ti-6Al-4V (R. Boyer, G. Welsch and E. W. Collings, Materials Properties Handbook: Titanium Alloys, ASM International, 1994: k = 6.7 W/m K, ρ = 4430 kg/m³, c = 526 J/kg K, β transus about 995 °C, solidus 1604 °C).
BibTeX
@misc{tripathy2026flashtemperature,
author = {Tripathy, Manisha},
title = {Flash Temperature Lab},
year = {2026},
howpublished = {\url{https://untetheredatom.com/tribology/flash-temperature}},
note = {Interactive web tool}
}