Experiment: coast down distances (rolling resistance) @ various tire pressures

[gsasquatch]

Maybe the higher pressures move the deflection more to the tread rather than the sidewall, and thus the results of more resistance at more pressure.
I've never gone beyond 50psi, for fear of a blowout, which would negate the advantages of over-inflating from an economic and environmental point of view, especially for me since 1 tire would be 10% of my annual fuel budget. This data makes it much harder to justify the risk of the higher tire pressures vs the efficiency gains.
I appreciate all the effort taken to get this data, but I wonder how did you come across so many tires?

[MetroMPG]

Originally Posted by gsasquatch

but I wonder how did you come across so many tires?

I have too many junk cars.

Five at the moment, though one is a parts car that will go to the recycler come nicer weather.

And I happen to have a couple of sets of extra wheels/tires.

[euromodder]

Originally Posted by gsasquatch

This data makes it much harder to justify the risk of the higher tire pressures vs the efficiency gains.

The unfortunate law of diminishing returns.

[darcane]

Originally Posted by MetroMPG

More inflation vs. rolling resistance data for the pile.

It matches the previous tests (meaning, rolling resistance decreases significantly up to a certain pressure, then plateaus).

Tire tested: Bridgestone Potenza RE92 LRR tires in 165/65R14 size (Honda Insight OEM tire, installed on Geo Metro electric car - ForkenSwift)




Weather: 25C, 8 km/h SW wind (test course ran NW/SE)

Methodology:

• tires were pumped up to 70 PSI, drove to test route (< .5 km), pressure adjusted
• car was reversed up a small hill (approx. 2 ft. elevation) and stopped at a marked point
• hand brake was used to stop - hydraulic brakes were not used at all during the test or on the short drive to the test area
• transmission in neutral, hand brake was released
• car rolled down short hill onto a flat run-out area
• where the car stopped, the road was marked
• pressure was adjusted (dropped 10 PSI)
• rinse & repeat

Results:



For kicks, see also: comparing RE92's @ 50 PSI to space saver spare tire donuts @ 70 psi - Fuel Economy, Hypermiling, EcoModding News and Forum - EcoModder.com - View Poll Results

Very interesting data here. One thing to note though, you have a very light car. The point of diminishing returns will likely be higher on heavier vehicles, like a full size truck or a Prius with all those batteries.

If you tried running only 50psi in the tires of a semi truck, you'd probably be highly disappointed in your fuel consumption.

[jime57]

Thanks for trying again, but I'm still not fully convinced.

What concerns me is that the conclusions run contrary to several world class hypermiles practices. 70-80psi is standard pressures for those trying to wring out the very best fuel economy. I don't think these folks are using those pressures because they like the ride Instead they have found the high pressures to be effective in delivering on another parameter - fuel economy.

I studied your chart and I see a couple things which don't look quite right. First, there is some inconsistency at the 60&70 psi level. In both cases, there were two significient outliers.

Second, not mentioned in the methology was warming up the transmission(I guess it has one) and tires. By starting cold and doing the tests in a reverse order of pressure, there is a bias injected in favor of the low pressures.

A question that remains unanswered it "Does roll down distance equate directly to fuel economy?" I'm not at all sure we know the answer to that, but hypermiles, by their actions, vote "no."

It is undeniable that tire pressures is a game of deminishing returns, but the real question is where the returns plateau. JMHO, but doesn't seem correct. All kinds of tests can be designed, some of which deliver misleading results. On the other hand of course, is 80 psi over 60 psi better to any significant degree for most folk. The answer is probably "no."

[MetroMPG]

Hi Jim -

I'll answer your questions as best I can:

Originally Posted by jimepting

What concerns me is that the conclusions run contrary to several world class hypermiles practices. 70-80psi is standard pressures for those trying to wring out the very best fuel economy.

I'd like to see their as-controlled-as-possible tests. I'm open to new evidence!

(Food for thought: lots of "elite" athletes perform strange pre-game rituals they wouldn't dare omit prior to an event. It's easy to see why a "higher must be better" theory might be adopted with respect to tire pressure.)

I studied your chart and I see a couple things which don't look quite right. First, there is some inconsistency at the 60&70 psi level. In both cases, there were two significient outliers.

No argument there. All I can say is I've done this test 4 times now on 3 different vehicles, and each experiment showed very diminished returns (to the point of changes disappearing in the testing variation) above the 50 to 60 PSI level. Could be limitations of the testing. Could be tire make as well. Maybe some tires continue to improve > 60 PSI, and I didn't happen to try those.

Second, not mentioned in the methology was warming up the transmission(I guess it has one) and tires. By starting cold and doing the tests in a reverse order of pressure, there is a bias injected in favor of the low pressures.

That's possible. I have no evidence on temperature changes, but judge for yourself whether transmission/tire warm-up is likely an issue or not based on this:

• The car was driven less than .5 km to the test area in summer temperatures at speeds less than ~30 km/h average (because I was using the hand brake to stop, to avoid engaging the front hydraulic brakes).
• Starting the test at elevated pressure means even less propensity for heat build-up on the way to the test area.
• The transmission was electrically driven, so no heat transfer from an I.C.E.
• The speed during the test was never much beyond about 5 km/h (whether coasting down hill or reversing back to the starting line).

Personally, I wonder if the transmission & tires ever got much above ambient temperatures throughout the experiment.

On the other hand of course, is 80 psi over 60 psi better to any significant degree for most folk. The answer is probably "no."

I'm with you there. That's what I tend to think, based on trying this 4 times in 3 different cars.

I've won a couple of competitive hypermiling events myself (placed 2nd in another... incidentally, the 118 MPG winner of that one ran @ 50 PSI), and don't bother going above 60. I'd wager the outcome of those events (or long-term extreme hypermiling efforts in the real world) depends far more on other factors than 60 vs. 80 PSI.

[3-Wheeler]

Originally Posted by jimepting

....I studied your chart and I see a couple things which don't look quite right. First, there is some inconsistency at the 60&70 psi level. In both cases, there were two significant outliers.

I think we are seeing the repeatability of the test itself. This result does not surprise me, after attempting to perform repeatable coast down tests for several years and coming up short based on my limited measurement techniques.

Originally Posted by jimepting

...Second, not mentioned in the methology was warming up the transmission(I guess it has one) and tires. By starting cold and doing the tests in a reverse order of pressure, there is a bias injected in favor of the low pressures.

I see what you are saying Jim, but I also think that because it takes time to change the air pressure in those tires between each test, that the transmission will not really get warm during each coast down test. There is too much time in between each trial for the heat to accumulate.

So in this case, it may be more desirable and consistent *not* to heat the transmission as it will just as cool for each coast down.

I suppose one could *randomize* the order in which each test is run in order for any *heating effect* to reveal itself in the data.

Originally Posted by jimepting

....A question that remains unanswered it "Does roll down distance equate directly to fuel economy?" I'm not at all sure we know the answer to that, but hypermiles, by their actions, vote "no."

My educated guess would be that coast down distance is proportional to over all fuel economy. Why?

When I do P&G on the way home sometimes, my fuel mileage is highly influenced by how far the car coasts with the engine off. When the ambient temperature gets into the 100's(F), it's amazing how much longer the car coasts on the same stretch of road compared to spring/fall/winter temps, and my gas mileage seems to reflect this.

Look at my seasonal gas log for evidence of this.

Just my observations.

Jim.

[jime57]

Lots of good thought provoking discussion. I guess if I had more confidence that roll down distances had a direct one-for-one correlation to fuel economy, then I'd also try the experiment again. But I'd use a slightly different methodology and try to keep everything at operating temperature. After all, that is where the real results are achieved anyway. But the bottom line on the theory part is that I have yet to see a theoretical development which traces the relationship back to Maxwell's equations(That's a standard Electrical Engineering joke)

Certainly there is a promounced plateau effect. I have seen this in fuel economy testing in an Echo, under very tightly controlled distances. In that case, there was insignificant improvement above 50 psi. But I think the Insight may well be a different animal. Special LRR tires, special low loss hubs, special lubricants, well very light weight at least, and any number of differences. I'm willing to conceed that nothing above 60psi makes much difference, but I think the Insight is a different animal.

[MetroMPG]

Good point - it may well be. I tested the OEM Insight tires, but with the EV, not with the Insight.

[mikeyjd]

Great Thread!