Desgeegee's Build Thread Ford Focus

[desgeegee]

This is the final result! I'm sorry for the bad quality of this photo but it's a several days it's has been raining all days and i haven't been able to take a good photo with a good light.
However the panel is quite invisible and so are the screws.

[desgeegee]

Yesterday i tried to do the Coast Down Test for Cd evaluation:
-7 test from 180kph* (about 112 mph) to 90/80Kph* (about 56/50 mph). *Speed from speedometer. ( 7 test, 4 in one direction and 3 in the opposite direction, because the trial road isn't very flat)
- I use elm 327 interface + Forscan to record all data (time, fuel, real speed)
- Several months ago I found an excel file that models the experiment for the evaluation of Cd and the resistance rolling coefficient, by inserting some characteristics of your car like mass, frontal area, etc plus the insertion of the actual speed that you find.

I edited this file because i don't like playing with two variables to minimize the error between the model speed and the actual speed, so i modeled the Crr with the Stuttgart rolling resistance coefficient which depends on Speed and Tire pressure.

The results are:
Cd= 0.272 ( i think it's quite reliable)

Error= 0.37 ( i think it's acceptable)



What do you think about it? Any suggestions?

After finish the diffuser, i'll try again of course!

[MeteorGray]

Is the error factor larger than the actual results?

[JulianEdgar]

I think it is impossible to accurately calculate Cd* by coastdown tests, without very sophisticated data gathering (including real time wind/yaw) and very good modelling (including all rotational inertias).

Given the high speeds you are using, it might be possible to measure some relative changes in drag. Do the tests again with windows up / windows down and see how much consistent difference in speeds you can record after 50 seconds.

(* let alone to three decimal places!!)

[Vman455]

Originally Posted by JulianEdgar

I think it is impossible to accurately calculate Cd* by coastdown tests, without very sophisticated data gathering (including real time wind/yaw) and very good modelling (including all rotational inertias).

Given the high speeds you are using, it might be possible to measure some relative changes in drag. Do the tests again with windows up / windows down and see how much consistent difference in speeds you can record after 50 seconds.

(* let alone to three decimal places!!)

After trying coast down testing several times this summer, I agree with all of this. It was just about impossible to get data consistent enough to show a change in drag with the windows up versus down, let alone try to calculate the cD with that spreadsheet (which I tried the first time. It came up with a radically different cR and higher cD with windows up than down!).

I wish there was anywhere here that I could do 100+ mph legally.

[desgeegee]

Originally Posted by MeteorGray

Is the error factor larger than the actual results?

Yes it is and this is the data that gives me doubts. If you see the table, this error factor depends mostly by the difference between Vactual and Vmodel of 30 and 50 seconds.

30s: I've 107,6-107,4-109,1-103,2-104,2-106,4-108,2
Vmodel: 107,9 kph
I think that the only way to fix this "problem" is to collect more data and
do more test.
50s:Here there aren't some data due the traffic that forced me to brake..

What about to consider the mode and not the average(statistic mean(?))?

I need to do again this test in a major proper way of course

[desgeegee]

Originally Posted by JulianEdgar

I think it is impossible to accurately calculate Cd* by coastdown tests, without very sophisticated data gathering (including real time wind/yaw) and very good modelling (including all rotational inertias).

Given the high speeds you are using, it might be possible to measure some relative changes in drag. Do the tests again with windows up / windows down and see how much consistent difference in speeds you can record after 50 seconds.

(* let alone to three decimal places!!)

Yes this test isn't intended to find the real Cd but to find a "good" method (or refine and fix it for a better estimate)
to estimate it and make relative comparisons with the various upgrade made
I will do the Windows Up/Down test. I promise

[JulianEdgar]

Originally Posted by desgeegee

Yes it is and this is the data that gives me doubts. If you see the table, this error factor depends mostly by the difference between Vactual and Vmodel of 30 and 50 seconds.

30s: I've 107,6-107,4-109,1-103,2-104,2-106,4-108,2
Vmodel: 107,9 kph
I think that the only way to fix this "problem" is to collect more data and
do more test.
50s:Here there aren't some data due the traffic that forced me to brake..

What about to consider the mode and not the average(statistic mean(?))?

I need to do again this test in a major proper way of course

You will be wasting your time unless you do a basic test of a change that does alter drag - eg windows up / windows down - and ensure you can measure it. Chasing imaginary changes is no fun.

[desgeegee]

Originally Posted by Vman455

After trying coast down testing several times this summer, I agree with all of this. It was just about impossible to get data consistent enough to show a change in drag with the windows up versus down, let alone try to calculate the cD with that spreadsheet (which I tried the first time. It came up with a radically different cR and higher cD with windows up than down!).

I wish there was anywhere here that I could do 100+ mph legally.

I think is quite impossible to calculate the Cr without all coefficients and data of tyre and tarmac.. i used the formula that at least considers pressure and speed but i know there is an approximation error