frontal area (A)

[MetroMPG]

This has come up in a few threads, decided to start a dedicated one.

E.G.: in Reason with me before I destroy a classic

Skewbe says:

I have actually conceived of a way to determine frontal area (probably been done before):
1. take a picture of your car from the very front, halfway up it's height from as far away and with as much zoom as you have (not digital) so that it fills the frame but you can still see the outline. Use a contrasting background if possible.
2. measure the width of the license plate
3. use image software and figure out how many pixels wide the license plate is.
4. use the image software to outline the car, make the car entirely black and everything else white. Save it into an easy to parse file format (i.e. 256 color uncompressed bmp).
5. slap a program together to count the number of black dots in the black and white image file. (note, some image software may be able to tell you the area selected in pixels)
6. multiply that number by the pixels per inch (license plate width pixels/actual width).

And Bob's yer uncle.

----

Responses to Skewbe from that thread, below...

---

Originally Posted by skewbe

1. take a picture of your car from the very front, halfway up it's height from as far away and with as much zoom as you have (not digital)

5. slap a program together to count the number of black dots in the black and white image file. (note, some image software may be able to tell you the area selected in pixels)

For those of us who are not code jockeys, I actually found a program that will do this for you. I'll dig it up again if someone wants it.

[mustngr]

Yo Metro,

I have an idear for your idear .

A small laser level would hug the body more tightly as long as it was square to the wall, no?

It would render it a two-person job but would definitely be more accurate than shadows. The beam offset on mine is only 3/16". The difference could be calculated if yer into it, but is insignificant enough to be ignored altogether.

[MetroMPG]

2 points on the digital photo/pixel counting method:

[MetroMPG]

Other ways of getting this info about your car...

[2TonJellyBean]

Drive through a big snowbank and measure the hole. ;-)

[trebuchet03]

So... Typically you can find the cD value for almost any car on the market... That hasn't been modified

For an unmodified car, you should be able to infer it from torque and gearing (that is, in 4th you're making X torque and in 3rd you're making X torque at a higher rpm). So you should be able to get a function of rpm and torque... Then use that function at different speeds (different rpm)... and I've lost my train of thought...

But in theory, you should be able to calculate cDA from something like that... Hrmm, you might need to know acceleration though... iono, it seems like more work for a less accurate result. But if you really want to infer it, you'll have to get a better derivation than that

[MetroMPG]

Originally Posted by 2TonJellyBean

Drive through a big snowbank and measure the hole. ;-)

[MetroMPG]

Here's a good illustration of the amount of error you'll be dealing with if you use a camera without enough optical zoom to measure frontal area

[oneinchsidehop]

Originally Posted by MetroMPG

Here's a good illustration of the amount of error you'll be dealing with if you use a camera without enough optical zoom to measure frontal area:


If I could get a crisply focused version of the right pic, I think it'd be good enough to use the outline-and-pixel-counting method.

There's an easier way. Just move back, way back and crop. The perspective in the center 1/3 of a 50mm lens is the same as, say a 100mm lens. It's an old portrait photographers' trick (people generally look better with a 135mm lens on a 35mm camera), if you don't have the right lens with you, just move back and crop. Using a higher pixel count should get you pretty darn close.

You would have to move back the same amount using a flash light or laser pointer to get results with the same level of distortion.

Now if you had two metros, and used one as a guide for the pointer and the other as a template to be measured you could decrease the distance a great deal because the beam is not originating from a single point, the beam (if you did it accurately enough) would be square with the template.

But I think you should drive through a snowbank... and post a video.

[skewbe]

An excellent illustration, and I like the use of binoculars More people have those than telephoto lenses certainly. Rifle scope might help too.

And another good tip there: "If you can see the rear tires in the photo, it's a sign the zoom isn't strong enough. "