Making Sense Of Oils And Lubrication

All fluids have internal friction, making them resist flowing. Viscosity is the measure of this flow resistance. Without viscosity, oil could instantly be squeezed out from between bearing surfaces.

Oils lose viscosity as their temperature rises. The reason for this is that fluids are held together by molecule-to-molecule attractive forces, but temperature is a measure of average molecular activity. The warmer the fluid, the more energy its molecules have, and the more easily they can overcome their mutual attractive forces. As a fluid cools, its molecules have less energy to drive them apart, allowing the attractive forces between molecules to impede their movement past one another.

Engine oils must lubricate adequately across the full range of temperatures in a warmed-up engine—from the relatively cool conditions in the crankshaft’s journal bearings and lower cylinder walls to the much-higher temperatures in the top piston ring groove, the upper cylinder wall, and the exhaust valve guide.

Ideally, we’d like each part of the engine to have its own separate oil, thicker in the engine’s high-temperature zones and thinner where conditions are cooler. But practicality requires one oil for the whole engine. As a result, oil is coolest and most viscous in the crank bearings, where local temperature is of the order of 180 degrees Fahrenheit, and hottest and becoming watery-thin in the 300–350-degree top ring grooves and the even hotter exhaust valve guides.

Once diligent workers like Standard Oil’s George Saybolt (1853–1924) had created standard methods for measuring viscosity at specific temperatures, it was found that oils from Pennsylvania-sourced crude often performed better in engines than other oils of the same nominal viscosity grade. Through much testing, this was traced to their slower rate of viscosity loss with temperature, allowing them to carry more load between hot parts without rapid wear.

RELATED: Have You Ever Hand-Cranked An Engine In The Dead Of Winter?

The present method of quantifying rate of viscosity loss with increasing temperature, called viscosity index (VI), was developed in 1929 by Ernest Dean and Garland Davis and adopted as the ASTM International (formerly known as American Society for Testing and Materials) D2270 standard. It now defines two temperatures, 40 and 100 degrees Celsius (104 and 212 degrees Fahrenheit), at which viscosity is measured. The slope of a line joining those measurements on a logarithmic graph of viscosity versus temperature is the viscosity index.

That slope describes how fast that particular oil loses viscosity as it heats up within the range of engine operating temperatures. The VI of a Pennsylvania-grade base oil (oil before additives are combined with it) was arbitrarily set at 100, and that of a Gulf Coast naphthenic base oil at zero. Viscosity Index is defined only within that 104-to-212-degree Fahrenheit range. Two oils of identical VI in that range, for example, can still have very different viscosities during engine cold starting in northern Minnesota winter. Low-temperature viscosity is defined by a separate standard: the oil’s pour point (the lowest temperature at which it can be poured; a bit colder and the oil solidifies).

Subsequently, oils of much-higher viscosity index were created with VIs up to 250. Why should one oil lose viscosity faster with temperature than another? While a lot of PhDs have been awarded in this area, a simplified view tells us that the difference is created by: 1) temperature-driven changes in the shapes of oil molecules; and 2) differences in the “associativeness” or degree of mutual attraction between molecules. If a certain class of oil molecules tend to become more compact at lower temperatures, they will contribute less viscosity at those temperatures than will oils whose molecules remain in the form of long straight chains whose friction ties many molecules together. If oil molecules are more attracted to one another, flow cannot take place without overcoming that attraction.

Irving Langmuir (1881–1957) introduced the idea of polarity in molecules. Many chemical bonds consist of two or more atoms sharing one or more outer electrons, and part of the deeply non-intuitive mystery of quantum physics (a mystery to me, anyway) is that such electrons may spend more time at one end of the resulting molecule than at the other. This causes the molecule, which is electrically neutral as a whole, to be slightly positive at one end and negative at the other. This is polarity. As you can imagine, this can cause molecules to attract one another in certain orientations. Such an attraction can cause more polar oil molecules to thicken faster at lower temperatures than do non-polar molecules (lack of polarity is a major reason why polyalphaolefin-based synthetic oils such as Mobil 1 can have outstandingly low pour points).

Knowing or suspecting that differences in the viscosity indexes of different oils arise from physical causes like the above, petroleum chemists realized they could use those same effects to create additives that would raise the VI of an oil to Pennsylvania-grade level or higher. What they did was to create long-chain hydrocarbon molecules that were somewhat soluble in lube oil and which assumed a compact low-viscosity form at low engine temperatures (rolled up into something ball-like or folded like a carpenter’s rule), but as molecular activity increased at higher parts temperatures, the resulting molecular pummeling would “unroll” such molecules progressively, slowing the loss of viscosity.

This is the basis of today’s multigrade oils—those with API service ratings such as 20w-50. To make a 20w-50 multigrade oil, you begin with a thin oil having a 20 rating at zero degrees Fahrenheit, and you add long-chain polymers to it to slow its loss of viscosity as temperature rises. The oil does not gain viscosity as it gets hotter, it just loses viscosity more slowly than would the base 20-weight oil. By the time that 20w oil has reached 212 degrees, its viscosity has fallen only to that of a straight 50 grade at that same temperature. Because its viscosity has fallen less, it can carry more load without being squeezed out from between moving parts.

Such multigrade engine oils were first marketed around 1950 as “all-weather” oils. I well remember the slow grinding of the 6-volt starter as my parents tried to start the family car in cold weather. Multigrade oils eased that task, allowing starters to turn engines faster, to start in weather that used to be “maybe” at best.

At first, many oil users, especially those operating heavy-duty diesels, were skeptical. Early VI improvers were fragile, their long chains breaking in higher temperatures or under conditions of high shear (gears, cam lobes, and valve tappets). This chain-breaking caused such oils to “fall out of grade” as they aged, losing hot viscosity as the long chains of the VI-improver additive were broken. Another fear was that multigrade oils, because they consisted primarily of a lower-viscosity base oil, would suffer more rapid evaporation loss from hot cylinder walls (which did happen in our race van in the 1970s, using a popular 20w-50 oil).

The problems were overcome by intensive development; even the diesel operators eventually accepted multigrade oils. Thanks to such oils, the hottest engine parts are now durably lubed by oil that has lost less viscosity than would a non-multigrade or “straight” oil.