Friction Power Loss at a Sliding Contact

Multiply the friction force by the speed it is dragged through and you have watts. That is all this page is arithmetically, and it is the number that decides whether most sliding parts live or die — because sliding contacts are overwhelmingly limited by temperature rather than by strength.

Consider the scale of it. A modest contact carrying 1500 N at 3 m/s with $\mu = 0.2$ dissipates 900 W. That is a domestic kettle element, and it is being delivered into a patch you could cover with a thumb, in a part that has no fan, no fins and no coolant. The heat has to leave by conduction into the surrounding metal, by whatever convection the installation happens to offer, and by whatever the lubricant carries away — and if the sum of those is less than what is going in, the temperature simply climbs until something gives.

What gives depends on the material. A polymer bushing softens, its clearance closes, the contact pressure rises, the friction rises, more heat goes in, and the process runs away in minutes. A grease-lubricated contact cooks its thickener and loses the oil. A metal pair loses the oxide film that was keeping it in the mild-wear regime, and transitions to adhesive wear and seizure. In every case the failure arrives long before anything reaches a strength limit, and this is exactly why plain-bearing catalogues are written around a PV limit — the product of contact pressure and sliding velocity — rather than around load alone. PV is this equation with the $\mu$ divided out and the load turned into a pressure, and its units are watts per unit area: a heat flux at the interface, dressed as a bearing rating.

Two feedback loops sit on top of the plain multiplication. The first is stabilising: $\mu$ generally FALLS as a contact heats, so the power input drops a little as the temperature rises. That is brake fade seen from the other side, and it is why the heat is somewhat self-limiting even though the temperature is not. The second is destabilising: in a lubricated contact, heat thins the oil, a thinner oil gives a thinner film, a thinner film means more asperity contact and higher friction, which makes more heat. A journal bearing with inadequate oil flow can walk itself up that loop and into the boundary regime without any change in load or speed.

So ask three questions of any answer this page gives. Which coefficient did you use — it should be the KINETIC one, since the contact is by definition moving. Where do the watts GO — name the path, and be honest that a bearing bolted into a still, insulated housing sheds far less than the same bearing on an open bench. And what temperature results — which is the calculation this page does not do, and the one that actually decides the design. A contact that makes 50 W and can shed 5 W is not a bearing. It is a heater with a service life.

One useful inversion: run the equation backwards from the watts the assembly can genuinely lose, and it hands you the load or the speed limit the thermal budget imposes. That is the honest way to size a dry bushing or a wear pad, and it is usually a much tighter constraint than anything the strength calculation produces.