Fuel Economy Improvement

Lyn, I do believe that day is here. I feel the consumer needs to expect or demand the automobile manufacturers to invest in new research & test methods for improving fuel effeceincy, similar to the demand on improving quality & safety. Paul

Yeah, suspicions abound.

As far as turbulence forces when passing tractor-trailers would be negligible, and not worth the time to study. Forces like this are what are typically factored into the margin of error when doing these studies. You cannot build any data based upon seemingly random events that you have no way of controlling.

RVZ717, my initial "fish-scale" testing indicates that these forces are not neglible, but some vehicles are more prone than others. Why? Paul

If we improve fuel efficiency, on the developed & quality proven engines, with an infrastructure that already exists, it would buy time for the development of new engines & fuels that would not leave us stranded somewhere or have costly repairs & upgrades, in the rush for new technology.

This is actually commonly done, especially in the field of engines, and is going as far as researching and testing weight reduction by replacing vehicle copper wire with lighter aluminium or changing to a higher voltage (less current means less copper = less cost and weight).

It is all about varying degrees of improving the vehicle efficiency and reducing its cost. The ongoing product development continues.

Jack of all Trades, I concur with your discussion & have witnessed such. There have been significant reduction in weight to reduce vehicle road load.

What I don't see or know about, is trying to do a similiar road load reductions with aerodynamic drag & tire rolling resistance with expressway traffic & surfaces.

I for one, have performed more expressway driving than any other type & would be delighted to get 3 or 5 more miles a gallon in that environment & under those road conditions. At slower speeds the result may be neglible, but for those of us that do commute, I believe there could be a significant improvement without dropping the speeds to 55 mph again. Paul

There have been significant reduction in weight to reduce vehicle road load.

Actually, in the US, similar model name cars have become heavier, not lighter. For example, a 1976 Accord weighed about 2000 lb, and the current one weighs about 3500 lb. This trend has been market driven. You can still buy a Civic, which is lighter and smaller than a current Accord (but heavier and larger than early Accords), but many people buy the Accord instead -- people like larger, more luxurious cars. SUVs are obscenely heavy and boxy, and sell well not because people are forced to buy them, but because people believe they want a huge heavy vehicle.

What I don't see or know about, is trying to do a similiar road load reductions with aerodynamic drag & tire rolling resistance with expressway traffic & surfaces.

Current tire rolling resistance varies in the range .006 to .014 (x vehicle weight). Low rolling resistance tires have been a research and marketing interest for decades. Priuses use very low rr tires, and such tires have less than half the rr of tires from a couple decades ago. Although you do not see this, the research has been ongoing.

Aerodynamic resistance is also low in vehicles for which there is a market demand for low aero resistance. (In the US, there are just two vehicles which are designed with fuel economy being a very high priority, the Prius and the Honda Insight -- and both have good streamlining.) In most cars, lip service is given to streamlining, again, because people do not really care enough about fuel economy to give up the styling that they like. The Prius was criticized for its appearance, and its appearance comes largely from its low Cd. The old (60's) Jaguar XKE looked great, (and many people thought it appeared "streamlined') but has an incredibly high Cd of about .40. The Prius is about .25.

I believe there could be a significant improvement without dropping the speeds to 55 mph again.

An average car of today has a Cd of about .32. That could be reduced to .20, without the car looking so strange that it would not sell. (This Daihatsu UFE is sub .2):

The Prius sells well when fuel prices are high, and not so sell when fuel prices are low. Anyone can buy a Prius and get 50 mpg. But not everyone does. The problem is not that the manufacturers cannot produce highly-efficient cars with low rolling resistance tires, good aerodynamics and highly efficient power trains. If that's what everyone wants, we'd all be driving Priuses. Fuel efficiency is not enough of a priority for most people -- just look at what most people drive.

"You cannot build any data based upon seemingly random events that you have no way of controlling."

Welcome to Earth. How has your first day here been so far?

OR..

That rule of scientific analysis was tossed out years ago, too restrictive to the para sciences, and now no longer applies to reality or the study of real life events or conditions.

So far earth has been nice... If I could get rid of all these pesky humans, It would be a much nicer stay.

My only point was that it would be a waste of time to study the effects of passing a semi truck, Not that It is not possible to do so. Anything can be measured, analyzed, evaluated etc etc...

What I meant to say was something like: you cannot build any meaningful data... or something of that nature.

Your feeling is that the mileage on a truck route & a highway populated by commuters [both at near capacity] would be the same [beyond the accuracy of the test to detect]?

there's surely wind tunnel data & other simulations, to check stability in these situations

tcmtech, is this Earth where I ended up this time. When I get lost in my own mind, I never know where I land.

My discussion was based on my observation, that while scientific analysis under laboratory conditions has merit, I feel there needs to be more real world testing, so that we actually achieve the "window sitcker" fuel ratings in the real world. Somehow, someway, I always come up short! Paul

P.S. I thought you enjoy this:

The Faithful Measurement Postulate

If an instrument is thrown into a definite state by a specific state of the system it is measuring, and if the same state of the world always produces the same state in the instrument, it will always be possible to infer the state of the world from the state of the instrument.*

*Reference Romano Harr'e "Great Scientific Experiments"

I feel there needs to be more real world testing, so that we actually achieve the "window sitcker" fuel ratings in the real world.

If you drive exactly as the EPA dyno operator driver drives, going up and down real grades equal to the test routine simulated grades, etc, you will get the same mileage. The EPA dyno tests were derived from real, observed, traffic conditions and produce results within a few percent of real results. I live in a place, and drive in a way, that produces figures that match the EPA figures. My new Civic and my older Accord both get the EPA numbers, my friend's Prius gets 50 mpg with me driving it, etc. The EPA site accepts data from the public, and when the sample size is large, the user mileage ratings and EPA ratings are very close.

To see the low utility of real world testing, just read TCMtechs review of his driving experience in a Prius. While most people get very close to the EPA numbers, TCMtech did much worse -- he can't drive, I guess. His "real world" test measures the driver, not the car. To measure the car, you need a dyno, the protocol, and all the recording equipment. Second best is a large sample size of actual drivers -- and that data is available on the EPA site, and tends to agree with the EPA test figures for many cars.

But for development, only dyno (and CFD, FEA, other computer models and simulation, etc.) makes sense. Cars already cost enough without having to build hundreds of protypes. Given a BSFC chart from an engine, a tire spec, a transmission spec, a body Cd and frontal area, and a gross weight, an engineer can get within a couple percent of the EPA test numbers, prior to actually building a prototype.

Manufacturers do loads of real world testing, but it is not a good substitute for real dyno testing, which is controllable and repeatable.

K Fry, thank you for the time & effort you have put into the discussions. Your technical insight & experience shows & is greatly appreciated. I agree with the technical apsects of your discussions & this is exactly what is being performed in the industry today.

I started this discussion by brainstorming on a Next Generation of Testing, that would lead to identification of improvements in fuel effeciency for specific environments & situations. The most common & where the most significant improvement I feel can be made, is in freeway traffic at 70 mph.

It appears to me that the Aerodynamic Drag or Cd value in many cases, seems to be derived from Wind Tunnel Testing or from a specific coeffecient of the road load polynominal derived from SAE J2263, which is a coastdown procedure that does no allow the vehicle to interact with passing or being passed by another vehicle. These are indeed the controllable & repeatable tests that are necassary as you so accuately described & every manufacturer performs them.

Where I feel the Wind Tunnel testing falls short, is that in many facilities the tires are not rotating, so it does not include wheel well turbulences & also the lack of underbody turbulences of a vortex shedding variety.

With SAE Procedure J2263, some take the road load polynomial coeffeicients as: 1 being aero, 1 being tires & 1 being chassis friction. I personnaly feel this is an oversimplified equation & erroneous interpretation. Since the coastdown runs are done in pairs, the effects of a head wind or tail wind are averaged or negated. Where is the validity of the aero?

I am extremely interested in you comments to this discussion so far. Paul

It is interesting to look at the EPA numbers and actual driver results for two of the biggest selling cars, the Toyota Camry and the Honda Accord. I picked the 2005 models to ensure that a relatively large number of people responded with actual figures.

You can compare cars side-by-side on the EPA site. (http://www.fueleconomy.gov/feg/sbs.htm ) Unfortunately, you cannot then provide a link to the comparison... only a link to the basic site. The comparison is very easy to do, however.

The cars (4 cyl, automatic) are given the same rating by the EPA: 21 city, 31 highway. Due to rounding differences, the combined figures differ by 1 mpg. 24 for the Honda, 25 for the Toyota. The averaged real world numbers are 26.4 for the Honda and 26.5 for the Toyota. In all respects the cars seem essentially identical, as would be expected, given their nearly identical specifications, aerodynamics, weight, etc. The range of "real" figures is also nearly identical: 16 to 35 for the Honda, 18 to 35 for the Toyota.

So it seems that the current EPA procedure does quite well, slightly under predicting combined performance. The range in reported mileage figures is to be expected, and is large, also to be expected, because different people have dramatically different driving patterns.

The coast down tests (even the older ones that averaged two way runs without benefit of on-board anemometers) get plenty close enough for the intended purpose. For the EPAs intended purpose of providing are performance yardstick, the coast down process and calculation is not an oversimplification: it provides essentially identical performance figures for cars that are known to be essentially identical. Even for manufacturer's use in body shape development, coast down tests have their place. The basic theory (that aero drag increases with the square of speed but that tire drag and final drive friction are nearly constants) adequately accounts for the three sources of road load, in close-enough proportion. Repeatable, accurate, mileage testing cannot be done on the road -- it has to be done on a dyno. To set the dyno load, you need to know the aero drag profile, and so you either have to do coast down tests, or accept the manufacturer's Cd numbers.

For real development work, coast down tests are not used, and wind tunnels are not often used. Where wind tunnels have been used for serious car work, they have virtually always been of the moving road type. But CFD has all but replaced wind tunnels for development work.

Where is the validity of the aero?

For the EPA tests, the calculated figures are close enough to real figures that the manufacturers are not complaining. The results are close enough that real world and EPA test results correspond well, and consumers are not complaining. On the other hand, for development work, calculated values from coast down tests are of little utility (unless you have a tiny shop like mine, in which case they -- and tufts on the car, etc. -- provide good information.)

The President already told us to check our tire pressure and get a tune up..............................that's all we need to know.

If I attach strong magnets to the front of my car, will they help to propel me closer to the car ahead of me? Unless it is Lyn"s Corvette.

Have a look at http://ecomodder.com/ There, I learned that wet pavement increases rolling resistance, among other fine points. Everybody from this forum should hang out there, long enough to realize that the format here is terrible in comparison. It is hard for me to believe we are a bunch of engineers, putting up with these inconveniences.

http://ecomodder.com/
the pics don't work very well

much more limited in the range of subject matter

wrong kind of community for most of us

An additional is just plain correctly using what we have and preventative maintenance goes a long way to large fuel savings. Incorrect inflation pressure of European vehicle tires wasted about 2 billion litres of fuel last year.

http://www.gizmag.com/under-inflated-tires-european-study/18473/

It all adds up.