Accurately measuring wheel circumfrence

[bowlofsalad]

Hello,

I've come across several techniques, thought of some of my own and so own for getting a very accurate measurement of a wheels circumference.

I came across https://www.youtube.com/watch?v=ck27TNqFJE0 this video. This seems like an ingenious method for making this measurement. I might do it two or three times and compare or add the numbers together and divide to be very certain. Seems like this might be useful information.

The only change to his technique I might do (I haven't actually tried his way yet) is I'd use a pair of scissors instead of drawing a line. Also, I'd be very sure I was accurately measuring the length of the painters tape.

 

[Hillhater]

Overcomplicated. !.. he could at least have used a metric tape !
But better still just do a normal "roll out" on the floor..
Put a mark on the side of the tire..
Match it up to a mark on the floor.
and then simply roll the wheel forwards until the tire mark is at the floor again ( no slipping) and put another matching mark on the floor.
Measure the distance between the marks .
Best done with the rider seated to get accuracy with the tire deflection under load.

 

[Gregory]

Best done with the rider seated to get accuracy with the tire deflection under load.

]

Agreed .
Why do they use three magnets for the speed sensor? Is it a noticeable improvement in accuracy or update rate?

 

[bowlofsalad]

Best done with the rider seated to get accuracy with the tire deflection under load.

]

Agreed .
Why do they use three magnets for the speed sensor? Is it a noticeable improvement in accuracy or update rate?

I think that perhaps at low speeds you might have a slightly more accurate speed or RPM reading with more poles or magnets rather than less.
The problem with the roll out concept, for me at least, is keeping things perfectly straight. That is why I was/am considering using some kind of guide to roll both wheels against to keep things straight.

 

[Ykick]

Last time I used the roll-out method it kinda bugged me about the wheel/tire at exactly the same point after full rotation? To help insure this position I placed a small bubble level inside the rim, marked it's exact position with pencil, used thin maksing tape to mark the sidewall to the floor position.

Rolled it one revolution, placed the bubble level in the marked position inside the rim and made my mark on the floor. Measure between marks on the floor and my speedo is now very, very accurate.

I suppose you might be able to tape the bubble level inside the rim but Pencil marks worked fine for me?

 

[hjns]

Good thread. Makes me want to go home and measure my wheel circumference accurately.

I will probably unroll my 5 meter tape on the floor. With my ~ 2070mm circumference, I should be able to have 2 rotations and measure the actual distance. Then dividing by two will decrease the error by a factor two as well.

 

[Rassy]

You guys are too anal. No speedometer should ever be assumed to be absolutely accurate, even on cars. Tire wear, speed, load, pressure, etc. will all affect the speedometer and odometer readings on bikes and cars, etc.

Having said that, I like the roll out method with the marks on the tire and floor. I just stick a piece of masking tape on the tire/floor and mark on the tape. It's quick and easy, and yes, use a tape marked in metric, and your final result will be just as accurate as someone that labors and worries about each mm would get.

 

[hjns]

Tire wear, speed, load, pressure, etc. will all affect the speedometer and odometer readings on bikes..

Let's see. Let's assume that the circumference of a "26 inch" new nobby tyre (660mm OD) is indeed 2070mm.
(check => Circumference = Pi (3.14) * Diameter (D) = 3.14*660mm = 2072mm)

Now the wear of the 1/2 inch nobs removes a maximum of 1 inch from the diameter, so the circumferences changes to 3.14*635mm = 1994mm, which is a difference of more than 75mm.

Assuming a distance of 2km, and a setting of 2070mm in the CA (966.1 rotations of a new tyre), with complete worn tyres, the CA will show the same 2km after only 1.926 km, which is 3.7% off.

 

[alan]

I lay out a tape measure on the street in front of my house, up to 50' feet long. Then I put my tire with the air valve immediately above the start of the tape, with the tire just to the side of the tape. Then I carefully roll the tire just to the side of the tape, counting the number of tire revolutions until I get near the end of the tape, and stop when the air valve is vertical, at the bottom of the tire again. (More accuracy would result if a piece of tape was put across the tire (radially, not circumferentially), and use the front (or back) of the tape to note exact rotations on the tape measure.) Note the length of the roll on the tape and divide by the number of revolutions.

Note that tire circumference stretch is virtually zero, so it will make no difference if you have weight on the bike, or whether the tire is spot on normal inflation. There will be more variance due to road surface smoothness, and that difference can be amplified by tire inflation and weight (whether the tire rolls over the top of surface irregularities, or bends into them).

There is one straight line segment of my commute that is three miles long. My odometer clocks exactly 3.00 miles on that stretch. When I accelerate towards a roadside radar speed sign (when no cagers are behind me being picked up by the radar), the indicated speed clicks up exactly with my spedo. That is, when my speedo goes from 19.9 MPH to 20 MPH, the sign goes from 19 to 20, and does the same on deceleration.

-- Alan

 

[itchynackers]

Blob of touch up paint on tire, roll forward, measure between edges of blobs on garage floor.

 

[www.recumbents.com]

Just use the valve stem for the "tire mark".

 

[Chalo]

Blob of touch up paint on tire, roll forward, measure between edges of blobs on garage floor.

I have used Liquid Paper to do this job before. It allows correct rollout measurement of a fully weighted tire.

Remember typewriters? Remember writing stuff down on paper?

 

[Kingfish]

I used a string. However I like the "blob of paint" idea better :wink: We can use anything for that, something biodegradable, water & flour, food coloring, spot of coffee, whatever.

Though Rassy is correct for issues that affect true OD readings when on the road, and our measurement is only valid for static testing. Now I have 2WD, and even though I have triple-measured each wheel (front is 26" and rear is 24"),, and they are both running dual-ply Hookworms, both CAs report different distances which becomes larger the longer I ride, naturally. Therefore as a rule, I take the average between the two computers and use that as my "reported" value, including the stats in my Sig.

I've observed that the error can magnify with change in tube pressure, large change in temperature (directly affecting pressure), road conditions (sandy, gravelly, wet, slick, paint, dirt, icy), and especially differences between the front and rear traction - each uniquely slipping a bit depending on the conditions and steerage. The front will turn more than the rear - which is a follower. However the front can slip in loose terrain or on an incline or on road paint. The rear can also break and slip if it gets mired. Generally I downgrade the power to the front by 10% to prevent that hub from breaking at a cold start as the mass shifts rearward. All of these little slips add up to error.

In the field, I'll often recalibrate when passing through a speedometer test; that's a great chance to dial in diameters more precisely because the tires are warm, the ground is generally level, and well... what else are we going to do for entertainment when were on the road? :lol:

Though for static measurement, the blob test seems easiest.
Touché, KF

 

[Drunkskunk]

Liquid paper and paint are too hard to clean up. I use ketchup, BBQ sauce, Salad dressing, or what ever is closest to my hand in the fridge door. It works just as well for the rollout method, but can be done on the kitchen or garage floor, and will wash off.

 

[cal3thousand]

Blob of touch up paint on tire, roll forward, measure between edges of blobs on garage floor.

Quick and dirty. I like this.

 

[Tommy L]

Circumference = PI x Diameter (PI = 3.141592)

This is Accurate as long as you get the Diameter accurate.

I've used a Tie Down Strap wrapped around the tire. The strap doesn't really stretch, so it works well


Tommy L sends......

 

[Dauntless]

If you're going to get sloppy like that, you'll never get your precise speed. While I agree with Rassy that exact just isn't important, since there's apparently people who just HAVE to have that 5th decimal, etc., you just have to realize there's only one way to get this just right, that is with an anemometer.

Just get a long threaded shaft bolt to replace the one in your gooseneck. The idea being there'll be a nut that you'll tighten to secure the handlebars, them the rest of the shaft will stick up for you to mount the anemometer. (Oh, try to not impale yourself on it if there's an accident.)

A properly linear anemometer has an error of less than 3% up to 60 mph (97 km/h) with maximum torque at 45 degrees to the wind flow. The three cup anemometer also had a more constant torque and responded more quickly to gusts than the four cup anemometer, so this is the better design.

What you'll do is have one of the cups painted in contrast to the others, i.e. one bright while the others are dark. You'll be counting the number of revolutions to divide by usually 2.5-3, depending on the particular anemometer.

Basically you're measuring the speed of the air rushing over you, much like a plane measuring air speed. If you get a fancy one not only will it do the calculating for you but you can get the wind chill at maybe 18mph, changes in humidity around the bike as you breathe hard, etc. http://www.toolexperts.com/airflow-anemometer-meter-wind-speed-chill-temperature.html?gclid=CJng2PHfwbcCFUQ6Qgodu0EAFQ

Wasn't that fun?

http://www.weatherwizkids.com/experiments-anemometer.htm

 

[bowlofsalad]

I don't know if caring about accuracy is really all that anal, I don't recall exactly what was the suggested default, but that is what I have now. I figure that isn't very exact, but considering bothering to measure the wheel, if a job is worth doing it is worth doing well.

Does one use this anemometer indoors? Would the wind have an effect on the reading?

The issue I have with any method that relates to physically rolling the bicycle to measure, even using BBQ sauce or string or whatever, seems like it could have some issues unless you had a way to guide both wheels (or all three) straight. I'd enjoy the idea of compensating for the variation in diameter while on the bike, but that sounds like a challenge to do while keeping the bike straight and all that.

I like the idea of using one of those sign speed detectors, but I have no idea where any are. I wonder how you find them.

 

[Dauntless]

Can't imagine a big enough indoors to use a anemometer indoors to test your bike speedometer. They're usually on the roof as part of a weather vane with a 2nd chicken or some such. The wind would absolutely affect readings, but no matter what direction it came from it would speed it up.

Radar speed detectors are most common around schools near me. There's been a trend toward universities having them on campus.

Last March 14th people were making a big deal about Pi day. (3.14). I said, "This is nothing compared to the Pi day they had back in 1592." The reality, however, is they weren't calling Pi 3.141592 yet. Boy, did THEY miss out.

 

[Hillhater]

you know... for $120 you can get a GPS bike computer that avoids all this measuring !

 

[Dauntless]

you know... for $120 you can get a GPS bike computer that avoids all this measuring !

Oh, but will it be PERFECTLY accurate? That's what the issue is.

The Earth is whatever 24,8xx miles in circumference. They divide it up into degrees which are just short of 70 miles each, or 60 nautical miles. One degree becomes 60 minutes, each of which is one nautical mile, 6,076 feet, then you break the minute up into 60 seconds/101.2666666667 feet. The Navies of the world have cables, leagues, etc., it can get precise.

Can you find him the GPS that won't miss a second, a cable, a league?

 

[itchynackers]

This GPS issue has been discussed ad nauseum before. The commercial handhelds/car units aren't accurate enough (I'm a professional surveyor/engineer). They don't have the proper data loaded to output an accurate distance (if that is what you want). If you want details, begin by reading here... http://www.ngs.noaa.gov/PUBS_LIB/gislis96.html ....then read some more.

 

[John in CR]

You can use a GPS to get accuracy to +/- 0.2% by comparing a straight road distance of a kilometer or 2 as reported by the GPS vs your bike's device.

Wouldn't the best way to get pretty extreme with accuracy be to use a straight road with mile markers and 1/10th markers, and then compare your bike's reported distance to the markers over say a 10 mile distance? Aren't those markers properly surveyed distances?

John

 

[Hillhater]

....and what are we going to use all this super accurate data for ?
does it matter if i am doing 19 mph or 21mph ?

 

[bowlofsalad]

....and what are we going to use all this super accurate data for ?
does it matter if i am doing 19 mph or 21mph ?

I've heard that argument lots of times used in various discussions. Does it really matter if a chef washes his hands extremely well? Does it really matter if a surgeons blade is perfectly accurate? Does it really matter if the Earth is perceived as flat or round? It's hard to say. But if all I really needed was my sense of speed and how much power I have and currently an using, why did anyone buy a cycleanaylst?