Rolling Resistance vs. Temperature
 

Has anyone ever seen any data for tire rolling resistance vs temperature? I've searched around and so far haven't been able to find anything.

From my own experience, I know that it goes up significantly as the temperature drops. For example, I typically back into my driveway (which slopes off towards teh street. The street is basically flat for about 100 ft or so and then slopes down slightly out to the first stop sign. Typically, I just start out in neutral in the driveway and just coast to the stop sign. I never get going very fast (3-4 mph tops). When it's really cold, however, (<20 deg F), I usually can't make it past the flat spot on the road and I have to start using the starter.

Another example is a section between stop signs I drive frequently. I do one pulse & glide for the whole section (~0.5 mi). When it's warm outside I only need to pulse up to ~30 mph to be able to coast the whole way. When it's really cold, it can take a pulse of ~40 mph to make it the same distance.

I watch my tire pressure and keep them pretty constant regardless of temperature, so I think I've cancelled out that effect. Also, I'm talking about low vehicle speeds so aero should be affecting things much.

Anyone ever seen hard data on this subject?

I don't have the source handy, but I saw a study of bicycle tires that said RRC doubles from 30C down to 0C.

I've noticed a similar effect on my coasting in the cold. I figured it was because things like bearings, fluids and the drivetrain in general (maybe the rear differential in your case) are all much tighter when its cold and are creating extra drag. Maybe the differntial fluid is slightly denser as well. Heat tends to make metals expand right. So mabye thats why the car doesn't roll as well. For example I've heard of people using dry ice to help press things into a motorcycle case. Never done it myself so I'm not sure if it works. Just a thought on why your truck might not roll as well.

I'm sure things like differentials, wheel bearings, etc, all play a role. I'm thinking tires probably are still the predominant source of rolling resistance, but I could be wrong.

Ya I'm not sure either. You could try heating your differntial/transmision along with all the other components that you already preheat and see if you notice a difference. Its not really that cold here at the moment so I cant really properly test it myself. Edit: [At least it would eliminate those factors as variables in the equation]

Using ecomodder's calculator with my car's specs, doubling the RRC drops the 35 mph mileage by 31%. That's without any oil viscosity changes, or air density changes, or the cold engine is more powerful effect. Sounds about right to me.

Hi Diesel_Dave,
See here.
The EPA temperature adjustment for tire rolling resistance:
So tire RR decreases about 22% at 0 compared to 68 degrees F. Which is going the wrong direction. The largest contributor to RR in modern tires is the stiffness in the steel belts, and that is almost unchanged for the range of temps that a tire sees. The change in the flexibility of the rubber at low temps is likely unimportant. In fact, as the rubber gets stiffer it flexes less so it should roll easier. I'd look at the viscosity of bearing and gear greases as the main culprit.
-mort

So that means they assume it's linear and the RR increases 1% for every 3 deg F you drop from 68 deg F.

That seems ballpark reasonable. Several flolks around here have a rule of thumb for FE vs. temp which is ~1% FE change per 2 deg F change. I think if you calculate air density changes you can ascribe ~1/2 that to aero.

This paper from '77 does talk about the changes to the rubber hysteresis, however, they don't give much data (see page 8):
http://www.edccorp.com/library/TechRefPdfs/EDC-1038.pdf


I also found a reference to an SAE paper from '80, but haven't found the full text yet:
http://papers.sae.org/800090/

I disagree.

As a winter bicyclist I'm keenly aware of a massive increase in r.r.. Bearing losses for motored vehicles are considered by SAE to be small enough to ignore in performance calcs (and their bearings have draggy lip seals); indeed when I spin a bike wheel in the cold it seems to spin nearly as well as ever (WARNING! Completely subjective!). I blame the tires themselves.

You say steel belts are the biggest contributor to r.r. when actually they are the biggest thing that reduces r.r.; steel springs back efficiently thus returning nearly all the energy it took to deform it vs rubber which due to hysteresis doesn't return as much when "undeforming"; thus the rubber is the culprit in absorbing energy when rolling.

I think the amount of flexion isn't changed much by temps (???) because the load on the tire is constant.

Oh, to have low r.r. steel belted bicycle and motorcycle tires!!! I think the super low r.r. tires used on solar racers are steel belted...

We know that tubeless tires have lower r.r. than tubed; more rubber = more r.r.

I once put some elastic bands in a fridge, and they became very stiff and hard to bend. One piece broke when it fell on the floor, which seemed very strange once the parts warmed up again.
Not all rubber will act the same, but tire rubber will not completely escape from this effect imho.

Also I think the stiffness increases exponentially with lower temps rather than linear. Otherwise the rubber band would not have broken like that.

I'd expect less than stunning performance from solar racers when temperatures are way below freezing anyway. Can't just blame the tires for that ;)

So I just repeated the test with some rubber bands, but they remained flexible ???
Ahhh, not all rubber is the same. Need the black vulcanized stuff that is not too flexible at room temperature, I guess.

You might be able to increase tire temperatures slightly with an Argon fill.

Argon is used as thermal and sound insulation between the layers of glass in higher end windows.
In a tire it should help keep the heat in the tires tread, since Argon offers lower thermal conductivity than air.

An Argon tire fill wont elevate temperatures to an unsafe level since there is a least one OEM that I know of who fills tires with Argon from the factory.
Lexus Argon fills their tires from the factory to reduce road noise, to at least help get the car off dealers lot.

A standard argon welding bottle regulator puts out right about 50psi, enough pressure to fill any car tire, any load range C and most LT truck tires.

I plan to argon fill my trailer tires to finish off my argon bottle when it gets low, not to reduce RR but to help fend off dry rot from oxygen permentation.

This would be true if the steel belt was a band of sheet metal. But it a woven belt, and flexing causes some of the wires to slide across other wires. through the weave, for which there is no rebounding force. That is pure hysteresis. I stand by my statement.
But don't try to slide the context out from under me. I'm talking only about steel belted radials. In other tires there are other contributors to RR. In non-belted tires (bias ply) squirm is the overwhelming source of RR.
-mort

http://deepblue.lib.umich.edu/bitstr...2B0?sequence=5

Here we are- an oldy but a goody. Don't worry; there's good stuff early on.

P.S. Here is a "little bit more"... :eek:

http://books.google.com/books?id=zcZ...stance&f=false

Pg 67 sez steel cords rule for fe. :thumbup:

I can add anecdotal information about viscosity and cold weather. When I start my Subie in the winter and let the clutch out (in neutral), the load it places on the engine is palpable; the revs drop, the computer compensates. The car even moves a bit, so I make sure the handbrake is set before letting the clutch out. For the sake of comparison on that engine (2.5L turbo flat four), when the AC comes on there is no discernable difference in load whatsoever; I just hear the clutch lock in.

That's just the oil in the transmission. No diffs are turning, no wheel bearings either.

I'd alert professional tire engineer (and contributor) CapriRacer to this thread.
.

There's a plot on pg. 5 showing the exponential dependence of rubber loss characteristics to temperature.

I seem to remember there's data in the NHTSA book on tires. I'm not near my copy at the moment so I'll post back when I've done the research.

But a couple of comments are in order:

Rubber hardly contributes at all to the stiffness of a tire. Stiffness is mostly inflation pressure related - at least as far as RR is concerned.

The steel belt isn't woven. It's 2 layers of parallel wires - one going one direction and one going the other. If you X-ray a tire, it may look like it is woven, but that's just an optical illusion.

Yes, a steel belt really changes the stiffness of a tire, but since we are talking passenger car tires where virtually every tire is a steel belted radial, the steel belt hardly makes any difference from tire to tire.

What really contributes to a tire's RR is the nature of the the tread compound. The technical term is hysteresis, and it is a measure of how much energy you put in to bend the compound, vs how much you get back out when you allow it to return to its original position. Even very stiff rubber compounds could consume more energy.

In fact, low RR tread compounds tend to be soft - just like grippy tread compounds. Long wearing compounds tend to be hard. So you can't tell if a tread compound is a LRR compound simply by measuring it's hardness.

So the rubber band experiment was not measuring the important property.

Bicycle tires have no steel belts and do have higher rolling resistance when cold.
So why would it not have that effect on car tires? As it seems to fit with the roll test observations?

I can add mine. Car tires cool off quicker than the engine and bearings. So leaving the car park in a workday evening after dinner in the cold, car still warm, sees bad free rolling in neutral, engine killed. Leaving in summer with a cold engine etc. got me rolling much further.
Same tires, 1985 Honda Civic, testing done in 2001, but no less true today.

My first link graphs cool-down and it really cools a lot in the first 10 min.

I know that this is still off topic for this particular discussion but I found this on another thread while looking into torque converter lock up. Brings me back to the cold causing more RR for the drivetrain possible even when in neutral. I still think this could be a factor in decreased coasting distances in cold weather.

[QUOTE=orange4boy;128030]Interesting find from an Amsoil ad, so grains of salt should be consumed:

Quote:
Over 40 percent of total energy loss in a transmission can be attributed to the act of pumping automatic transmission fluid (ATF) to the transmissions components during transfer of power. The mechanical efficiency of an automatic transmission is directly affected by the viscosity of the transmission fluid. When temperatures drop and cause ATF to thicken, transmission efficiency and power decline. Industry tests on torque loss demonstrate that a temperature increase of 150 degrees F increases transmission efficiency by up to 37.5 percent. In other words, as viscosity decreases, transmission efficiency increases.

This means to two things for this winter:

1) Synthetic ATF
2) Transmission preheater.


Quote:
While the average conventional transmission fluid has a Brookfield viscosity of 15,000 cP, Amsoil synthetic Automatic Transmission Fluid far exceeds the minimum standard with a Brookfield viscosity of 7,454 cP ensuring maximum transmission efficiency and power even during the coldest winter temperatures.