A few posts ago someone said the oil thickens as it heats up. My understanding of multi-grade oil is that 20w50 will pour/flow like 20w viscosity oil when cold and maintain 50w viscosity when hot.

The clearances racing engines run depends on a lot of things. One is the heat range the engine will be running in. I Formula One cars, that is quite hot; hotter than the average street car will tolerate.

The second is how much various parts expand. It is possible that those drag racing engines get hot enough that the moving parts will expand in their bores and bearings sufficiently to bind if the tolerances are tight to begin with. Again, it depends on the engines' heat range; since drag engines don't even have radiators, I expect they run very hot. Or not, depending. I understand that top fuel cars running on nitromethane can actually run quite cool because their fuel absorbs so much heat when it vaporizes.

Running pre-heated oil and coolant through the engine would indeed expand the clearances to acceptable size, rather than binding the engine up at the start.

It is also possible that cool oil simply won't flow through tight tolerances in a racing engine. And, as fastone371 points out, the pressure spike of a cold start and/or revving a cold engine may cause problems, and not just with the engine itself. I read an article by Pat Bedard, some years ago, when he was racing a Ferrari BB at Daytona. He said the engine had to be warmed up slowly or "the oil would pop the coolers." That's some pressure spike...

Finally, racing cars use straight distilled water in their cooling systems. This is partly because glycol is slippery when dropped on pavement, and because glycol actually has a lower boiling point than plain old water because it does not transfer heat quite as readily.

Turbo Magazine tested water, a 50/50 mix of glycol and water and several other products like Red Line Oil's "Water Wetter", DEI's "Radiator Relief," Hy-Per Lube's "Super Coolant" and Justice Brothers "Radiator Cooler" for their relative heat transferance properties and how hot the coolant got under the same test conditions. As it turned out, the glycol/water mixture ran hottest and the other products were about 5% to 8% cooler than straight water. Read the full report here -- http://www.turbomagazine.com/features/0703_turp_cooling_system_additives/index.html

That may not sound like a whole heck of a lot, but in Formula One, where everything is a matter of millimeters, a 5% reduction in heat can mean a 5% smaller radiator and a smidgeon less drag.

One final note: a cooling system pressurized to 3.75 bar (gage) is running at about 54 to 55 pounds per square inch -- 3.75 x 14.5 psi of stanard air pressure. That's much higher than the average automobile radiator system, which uses a radiator cap of only 12 psi (gage) or 0.87 bar (gage), which raises the boil-over temperature to 230°-250°F

There seem to be a few misconceptions here. The first is that multi-viscosity oil increases in viscosity with temperature. It does not. It simply loses viscosity at a slower rate with temperature. The curve is relatively flat.

Another misconception is that engines get tighter (less clearance) as they heat up. Although some of the parts expand at different rates, all materials expand with temperature and therefore the clearances also tend to increase with temperature (the exception would be a material with a high rate of thermal expansion that is inside a different material of a lower rate of thermal expansion). An inside radius will increase just the same as an outside radius. It is a myth that holes get smaller as they heat up.

Finally, although oils are measured and sold based on viscosity, it is a property called lubricity that determines the lubricating value of an oil rather than viscosity alone. Lubricity is essentially the ability of the oil to cling to the metal under shear forces and tends to be the opposite of surface tension, but it is not that simple. Like tack or stickyness, it is a property that somewhat defies scientific definition and must be measured empirically.

fastone371 - 18 October 2009 03:09 PM

Do F1 engines actually have LESS clearance than O.E.M. engines? Typically in drag race engines we run LARGER operating clearances although they are very precise. A typical drag engine has more bearing clearance, more piston to cylinder wall clearance, rod side clearance, wider valve lash, etc. The only tighter clearances I can think of would be ring end gap & piston to valve clearance. I think the reason for preheating the oil is to stabilize engine dimensions and to prevent a pressure spike on "cold" start-up.

I too am curious as to what type of tolerances they run, but they have longevity to factor in, so i think they would run a tighter clearance than in a drag application.
I once spent the weekend at a Dallas NHRA event with full access to Force's crew/pits and trailers a bunch of years back. Loose clearances in top fuel/FC engines is an understatement! I don't think they ever put more than 1 run on them, no way they'd survive. But the idea is to cut down on rotational friction power loss.
My own particular muscle car "play toy" came with a 2 year only production run engine that while the block casting did have its differences, was basically identical bore/stroke etc to the more common engine. The engineers did loosen up the tolerances on the crank and rods, then upped the oil pump from a 60 to an 80 psi unit. Valve cover has a sticker noting the owner to expect excessive oil consumption because of it.
I also agree with the above about pre-heating the engines prior to starting them. I think they have different materials in critical areas that might be in an interference situation until brought up to operating temps.

CHEMENG - 21 October 2009 01:00 AM

There seem to be a few misconceptions here. The first is that multi-viscosity oil increases in viscosity with temperature. It does not. It simply loses viscosity at a slower rate with temperature. The curve is relatively flat.

Another misconception is that engines get tighter (less clearance) as they heat up. Although some of the parts expand at different rates, all materials expand with temperature and therefore the clearances also tend to increase with temperature (the exception would be a material with a high rate of thermal expansion that is inside a different material of a lower rate of thermal expansion). An inside radius will increase just the same as an outside radius. It is a myth that holes get smaller as they heat up.

Finally, although oils are measured and sold based on viscosity, it is a property called lubricity that determines the lubricating value of an oil rather than viscosity alone. Lubricity is essentially the ability of the oil to cling to the metal under shear forces and tends to be the opposite of surface tension, but it is not that simple. Like tack or stickyness, it is a property that somewhat defies scientific definition and must be measured empirically.

I have to disagree with a couple of your points. In the 1st place, multi-viscosity oil certainly do gain viscosity as they heat up. That's why they are called multi-viscosity, or more commonly, multi-grade. Take 2 engines and put straight 30 in one and 5W-30 in the other and try starting them on a -20 morning and you will quickly spot the difference. Yet they both become 30W as they heat up, but the multi-grade will do so at a higher temperature. I have stacks of oil sample reports in my office that show that the 15-40 oil our company uses in our trucks and equipment is actually 40 weight at operating temperature. The 15 is the viscosity at what is called "standard temperature". Used for testing purposes. 69F, if I remember right. All oil thicken up as the temperature drops below standard, but multi-grades, especially synthetics, do so more slowly then straight grades. Multi-grades are simply tailored to do the job better.
Engine clearances certainly do tighten up as the engine heats up. For 2 reasons. One is the difference in materials. Pistons expand more the the cylinders or rings. In large part because they run much hotter and are of different densities. Bearing inserts expand at a different rate then the crank journal or rod and close the gap. The crank journal will expand at a different rate then the rod because of temperature differences. You also have similar or identical metals expanding towards each other, as in the valve train. An expanding cam lobe narrows the gap between it and the expanding cam follower and the expanding valve stem. Steel gears do the same thing. They expand into each other and close the gap. Temperatures within an engine are not uniform, so expansion, even with identical metals, is not uniform. Dissimilar materials add to the effect.

Cat-Man-Do - 13 November 2009 10:49 PM

CHEMENG - 21 October 2009 01:00 AM

There seem to be a few misconceptions here. The first is that multi-viscosity oil increases in viscosity with temperature. It does not. It simply loses viscosity at a slower rate with temperature. The curve is relatively flat.

Another misconception is that engines get tighter (less clearance) as they heat up. Although some of the parts expand at different rates, all materials expand with temperature and therefore the clearances also tend to increase with temperature (the exception would be a material with a high rate of thermal expansion that is inside a different material of a lower rate of thermal expansion). An inside radius will increase just the same as an outside radius. It is a myth that holes get smaller as they heat up.

Finally, although oils are measured and sold based on viscosity, it is a property called lubricity that determines the lubricating value of an oil rather than viscosity alone. Lubricity is essentially the ability of the oil to cling to the metal under shear forces and tends to be the opposite of surface tension, but it is not that simple. Like tack or stickyness, it is a property that somewhat defies scientific definition and must be measured empirically.

I have to disagree with a couple of your points. In the 1st place, multi-viscosity oil certainly do gain viscosity as they heat up. That's why they are called multi-viscosity, or more commonly, multi-grade. Take 2 engines and put straight 30 in one and 5W-30 in the other and try starting them on a -20 morning and you will quickly spot the difference. Yet they both become 30W as they heat up, but the multi-grade will do so at a higher temperature. I have stacks of oil sample reports in my office that show that the 15-40 oil our company uses in our trucks and equipment is actually 40 weight at operating temperature. The 15 is the viscosity at what is called "standard temperature". Used for testing purposes. 69F, if I remember right. All oil thicken up as the temperature drops below standard, but multi-grades, especially synthetics, do so more slowly then straight grades. Multi-grades are simply tailored to do the job better.
Engine clearances certainly do tighten up as the engine heats up. For 2 reasons. One is the difference in materials. Pistons expand more the the cylinders or rings. In large part because they run much hotter and are of different densities. Bearing inserts expand at a different rate then the crank journal or rod and close the gap. The crank journal will expand at a different rate then the rod because of temperature differences. You also have similar or identical metals expanding towards each other, as in the valve train. An expanding cam lobe narrows the gap between it and the expanding cam follower and the expanding valve stem. Steel gears do the same thing. They expand into each other and close the gap. Temperatures within an engine are not uniform, so expansion, even with identical metals, is not uniform. Dissimilar materials add to the effect.

Oils do not thicken nor gain viscosity as they heat up; their volume expands, but their viscosity lessens (common sense tells us that as a given mass expands, its density lessens, and viscosity is kind of a measure of density):

"Multi viscosity oils work like this: Polymers are added to a light base (5W, 10W, 20W), which prevent the oil from thinning as much as it warms up. At cold temperatures the polymers are coiled up and allow the oil to flow as their low numbers indicate. As the oil warms up the polymers begin to unwind into long chains that prevent the oil from thinning as much as it normally would. The result is that at 100 degrees C the oil has thinned only as much as the higher viscosity number indicates. Another way of looking at multi-vis oils is to think of a 20W-50 as a 20 weight oil that will not thin more than a 50 weight would when hot."
http://www.repairfaq.org/filipg/AUTO/F_oil_facts.html#OILFACTS_004

wilmywood8455 - 14 November 2009 06:43 AM

Oils do not thicken nor gain viscosity as they heat up; their volume expands, but their viscosity lessens (common sense tells us that as a given mass expands, its density lessens, and viscosity is kind of a measure of density):

"Multi viscosity oils work like this: Polymers are added to a light base (5W, 10W, 20W), which prevent the oil from thinning as much as it warms up. At cold temperatures the polymers are coiled up and allow the oil to flow as their low numbers indicate. As the oil warms up the polymers begin to unwind into long chains that prevent the oil from thinning as much as it normally would. The result is that at 100 degrees C the oil has thinned only as much as the higher viscosity number indicates. Another way of looking at multi-vis oils is to think of a 20W-50 as a 20 weight oil that will not thin more than a 50 weight would when hot."
http://www.repairfaq.org/filipg/AUTO/F_oil_facts.html#OILFACTS_004

Based on my own experience I have doubts about some of the claims made by the author of the article you quote. I speak from experience and not from any research on my part. You've made me curious though. I deal with Castrol's formulation engineers on occasion, so hopefully I'll have time on Monday to get their take on this. I'll get back to you.

Cat-Man-Do - 13 November 2009 10:49 PM

I have to disagree with a couple of your points. In the 1st place, multi-viscosity oil certainly do gain viscosity as they heat up. That's why they are called multi-viscosity, or more commonly, multi-grade. Take 2 engines and put straight 30 in one and 5W-30 in the other and try starting them on a -20 morning and you will quickly spot the difference. Yet they both become 30W as they heat up, but the multi-grade will do so at a higher temperature. I have stacks of oil sample reports in my office that show that the 15-40 oil our company uses in our trucks and equipment is actually 40 weight at operating temperature. The 15 is the viscosity at what is called "standard temperature". Used for testing purposes. 69F, if I remember right. All oil thicken up as the temperature drops below standard, but multi-grades, especially synthetics, do so more slowly then straight grades. Multi-grades are simply tailored to do the job better.
Engine clearances certainly do tighten up as the engine heats up. For 2 reasons. One is the difference in materials. Pistons expand more the the cylinders or rings. In large part because they run much hotter and are of different densities. Bearing inserts expand at a different rate then the crank journal or rod and close the gap. The crank journal will expand at a different rate then the rod because of temperature differences. You also have similar or identical metals expanding towards each other, as in the valve train. An expanding cam lobe narrows the gap between it and the expanding cam follower and the expanding valve stem. Steel gears do the same thing. They expand into each other and close the gap. Temperatures within an engine are not uniform, so expansion, even with identical metals, is not uniform. Dissimilar materials add to the effect.

I can find no reference to any liquid material whatsoever that undergoes an increase in viscosity with temperature and there are no lubricating fluids that I am familiar with that exhibit such behavior. Viscosity loss with temperature is explained by the reduction in intermolecular attraction as the molecules become more energetic (gases do just the opposite because their behavior is governed by the kinetic theory of gases – their viscosity increases with temperature!). I will explain oil behavior as follows:

First, the “weight” (W) of a motor oil is not true viscosity. It is a measure of the time required for a standard weight of oil to flow through a standard orifice by gravity at a standard temperature. True viscosity is defined as shear force per unit area divided by the shear rate (velocity divided by distance), and has the units of mass per distance-time. The standard unit is the Poise defined as 1 gm/cm-sec. The more common unit is the centipoise which is 1/100 of a Poise. If you imagine a flat plate that is separated by a film of oil from a moving plate, then the shear rate is the velocity of the moving plate divided by the distance. Increasing the speed of the moving plate or decreasing the distance between the plates increases the shear rate. The force is the force per unit area (also referred to as shear stress) required to move the plate.

If you measure the viscosity of a 40W oil in a standard measurement device, the time to clear the standard weight will be 4 times greater than that of a 10W oil. Here is the tricky part. If the oils are heated, EACH will clear the device in a shorter time. If you then measure a 10W-40 oil at the standard temperature it will flow at the same rate as a 10W. After heating the 10W-40 to the higher test temperature, it will flow FASTER but, not as fast as the 10W. Instead, it will flow at the rate of the 40W at the higher temperature. All of the oils flow faster at higher temperature because they all have less viscosity. The straight weight oils lose more viscosity than the multi-vis at the higher temperature, but they ALL flow faster. To be completely clear, the 40W oil flows faster at the higher test temperature than the 10W at the lower test temperature.


Thermal expansion can be rather complex when different materials are involved, but the general rule still applies that heat causes every molecule to expand. Holes get bigger when they heat up even if something inside the hole is growing faster. There is a fair amount of variability in thermal expansion with different steel alloys and different heat treatments due to differences in grain structure. Aluminum alloys typically have about twice the rate of thermal expansion of steel. Aluminum pistons will always grow faster than a steel sleeve, but titanium is just the opposite. Besides being a very tough metal, titanium alloys have very low rates of thermal expansion.

I am fascinated by your statement about gears and other parts moving into each other as they heat up. I would think that should depend on the construction of the case or block. If a transmission has an aluminum case, then assuming reasonable heat distribution, the shaft centers should move away from each other faster than the gears would grow into each other when heated, and the clearance should increase. Maybe that is a good reason not to drive through deep water and cool off the case too fast.

I have another thought on the topic of oil. A conventional engine in a street car uses oil pressure with fluid bearings because they can tolerate high loading, get low wear and good longevity. The penalty is more friction and lost power.

I understand that racing engines use roller bearings that have lower friction, but wear out much sooner. Is this true? If they are using roller bearings, then the relationship between clearance, temperature, and oil viscosity can be very critical. I have read that under certain conditions a ball or roller bearing can slide rather than roll and then develop an imperfection or flat spot that will lead to premature wear and failure. I don’t remember the details, but I think it had something to do with temperature and bearing loading. The author was making the case that oil is typically better for a bearing than grease. This may have something to do with the warm up issue. I believe this is a bigger problem for ball bearings than cylindrical roller bearings.

CHEMENG - 14 November 2009 02:50 PM

I have another thought on the topic of oil. A conventional engine in a street car uses oil pressure with fluid bearings because they can tolerate high loading, get low wear and good longevity. The penalty is more friction and lost power.

I understand that racing engines use roller bearings that have lower friction, but wear out much sooner. Is this true? If they are using roller bearings, then the relationship between clearance, temperature, and oil viscosity can be very critical. I have read that under certain conditions a ball or roller bearing can slide rather than roll and then develop an imperfection or flat spot that will lead to premature wear and failure. I don’t remember the details, but I think it had something to do with temperature and bearing loading. The author was making the case that oil is typically better for a bearing than grease. This may have something to do with the warm up issue. I believe this is a bigger problem for ball bearings than cylindrical roller bearings.

Not 100% certain this is the case, but the only engine rod or main bearings in a racing application I am aware of are in motorcycle engines.