How accurate is your odometer??

I have not conducted a scientific study but I casually observed that my GPS shows my speed at 59 mph while my speedometer displays 60 mph. This difference was noticed on several different days over a period of several miles. This is a 1.7% overstatement in actual speed. I imagine that my odometer would reflect this error too. You may wonder if my GPS is accurate. Well, I also noticed that the speed limit signs which display your speed using radar, seem to match my GPS speed rather than my speedometer. Incidentally my tires are properly inflated (on the high side). I don't know if this is associated with the fuel economy error that we have observed but it is interesting to note anyway.

 

The two gauges don't have the same errors. The speedometer reads high due to industry standards and some government regulations (some state laws in the US, some UN trade regulations in Europe and Asia). The odometer reads much more accurately, due to past class action lawsuits about warranty fraud (I received the legal notices for two older cars of different brands).

With about 20,000 miles on my OEM tires, my odometer read about 0.2% low.

 

Mine is pretty much spot on. There is very little difference between GPS distance and trip recorder distance. Easily within 1%. Although this could mean that after 100,000 miles, you could be 1,000 out.

Interestingly, the average speed indicated by the Prius is a lot more accurate than the speedo. If you are doing a constant speed (eg. on freeway/motorway), reset the trip meter and hold the speed constant on the speedo. The Av speed on the trip meter will read lower and closer to your actual speed. If your satnav has an average speed feature on its trip recorders, you can compare it with the Prius average speed over a trip.

 

Ken from Japan pointed the above out some time ago. I tried this out with cruise set at 60 mph. I reset Trip B and monitored Average MPH and Odometer. It took me exactly one hour to travel 60 miles (both on Odometer and Trip Miles) and the Average MPH was dead on 60.

There are many articles on here about the speedometer reading artificially high due to certain requirements but that the car odometer should be correct, or at least very close. At 60 MPH on the speedometer my ScanGauge reads 58 MPH. Also reading MG2 RPM on SGII at 60 MPH on level ground and calculating wheel speed through the HSD gear ratios this calculates out to 58.6 MPH. Even though the speedometer reads high it does appear that the car knows the true wheel speed. Of course changing out the tire sizes from spec will throw this off.

 

That's about as good as you can realistically expect from a car odometer. The tread has 0.25 inch of usable rubber, so the nominal 12.5 inch tire radius changes by 0.25 inch as the tire wears from new to worn out. That's a 2% change

 

Loaded with a car, and a contact path deformed to fit the road, the effective rolling circumference changes less than that. But I'm not able to quantity the change with real figures.

 

Then how about explaining the phenomena?

Now that you said that, I have an idea why there would be a tendency for that, but would like to hear your version of why.

Thanks for the thought provoking comment.:cheer2:

 

I have checked my odometer against a 3-mile marker on the highway to work and it is correct against this also.

 

That does beg the question: what degree of accuracy was used when installing mile markers?

 

What else would you like to question? Any other idea of how to measure distance and the odometer.

All of that is close enough for me to say that my odometer is accurate enough for my concerns.

 

No offense intended, its just that the conversation had already started to run down the direction of figuring out the impact of tire wear, so it seemed natural to start questioning other aspects of the measurement.

 

On long boring trips, I have found a very high level of consistency on mile markers installed for speedometer checks in different states (California and Arizona). Unless different states all have exactly the same error, they are within 1%. i.e they give the same reading over a 10 mile run within less than 0.1 mile.

I have also found regular highway mile markers to be very consistent, better than 1% variation(California and Arizona), although there is the rare odd one that isn't. If you are going to use regular highway mile markers, compare 3 or more sections of road (preferable different highways) until you find 3 that agree. In my experience, that will usually be the first three you try. I suspect most other states are accurate too, I have never checked them though.

In California, if a road has been realigned, it is indicated on the mile marker. Use realigned markers with caution unless you understand the process.
2003 Traffic Volumes on the California State Highway System

 

OK, sorry, I see where you are coming from with that question.

I can say that I have checked every car that I have owned against these mile markers over the years, three one mile markers. I always check the car odometer while checking the timing between the mile markers to determine the +/- accuracy of the speedometer. Every car odometer has checked dead on with the mile marker so I would have to believe they are fairly accurate.

That is back when I was commuting to work at the posted speed limit +10 mph and needed to know my true speed so to avoid the local speed traps

 

I don't have concrete reasons why, only some simple calculations and experiments with rolling radius vs tire pressure that don't match up to published revs per mile or real odometer results.

The explanation from others the seemed to fit best was that rolling circumference is strongly linked to the length of the steel belts just under the tread. This length does not change significantly with tire pressure or wear. A significant portion of the radius change with pressure and wear gets lost in sidewall squirm as the nearly fixed tread length keeps control.

What is your idea?

For distance reference, I like road mileposts because they are based on the original road survey. But there are problems, such as posts offset for installation convenience, length changes due to grade realignments (some marked, some not), and routing mismatches on divided highways. For these reasons it is necessary to check many mileposts for patterns, discarding inconsistent sections.

 

I checked mine with several GPS units over a long measured course. It's well within 1-percent, which is well within the delta caused by tire wear.
As a matter of fact.....mine is pretty much dead on with 21K on the tires.
YMMV.

I thought that speed and odometer readings are usually made from the same measurement. I wouldn't presume to know how the folks in Aichi do it, but on my bikes and my GMs I thought I read where one VSS sensor supplies both.
There's usually some form of device that picks off something like a shaft rotation somewhere in the drive train, and converts it into VSS, or Vehicle Speed. Different vehicles use different sensors, but I was under the perhaps false impression that this is usually converted into BOTH speed and distance traveled.
In fact, for my bikes, you can purchase a VSS modifier that will allow the driver (rider, in this case) to directly modify the VSS pulse train (by up to something like 10-percent, IIRC) to allow for some tire and pulley/sprocket modifications. For these vehicles, one each Harley and Kawasaki, I know for a fact that it affects both speedometer and odometer.
Like I said.....I'm suuuuuure that the slide rule drivers in Aichi do it differently, and I'm also sure that somebody on the forum has a fuzzy logic diagram that explains how and why.

I put the giveacrap switch in the "don't" position, since mine is pretty much dead on as measured several times over a long (over 50-miles to allow for GIS errors) course.

NOW......what I can't figure out is why my MDF is consistently 3+ percent off when it guestimates my MPG.....but hey, I know how to calculate it at the pump, so that's no big deal either.

 

I think you are close to the answer with the steel belts idea, a longer explanation that I haven't seen anyplace else takes into consideration the properties of a bending beam and it's neutral axis.

When you bend a beam, a common 2x4 for instance, the fibers at the outer diameter get stretched and the fibers on the inner diameter get compressed. Someplace in between those two extremes, the fibers don't get stretched or compressed. That layer is called the neutral axis. On a 2x4 the neutral axis is in the center of the 2x4 becuse the 2x4 is symmetrical about it's neutral axis in geometry and material properties.
[ame="http://en.wikipedia.org/wiki/Neutral_axis"]Neutral axis - Wikipedia, the free encyclopedia[/ame]

The hoop portion of a tire however isn't symmetrical about it's neutral axis in it's geometry or properties. The inner layer of the tire is a thin layer of rubber, the next layers out are the cords or belts of the tire, frequently steel on passenger car tires. The outer layer is the rubber tread, most of which is designed into blocks. Because the blocks have air spaces between them, they don't contribute much to bending stiffness. My SWAG is that the neutral axis is near the a outer steel belt.

What happens then is that as the tire tread makes contact with the road, that portion of the hoop section of the tires is deformed into a straight beam. When that happens, the neutral axis stays the same length by definition, the inner surface gets stretched and the outer surface gets compressed. The car only moves as far as the outer surface of the tire and since the length of the part of the tire on the ground is compressed, the car moves less distance than indicated by the diameter of the tire.

There are some other factors at play, but I believe that is the biggest one.:rockon:

 

Since the steel belts won't change dimensions under pressure as much as the surrounding rubber, it would seem that the neutral axis should be drawn closer to these belts than it would fall if the bulk material was uniform. So these two explanations end up with similar predictions.