Standing start calculation

Andy_Casey

I recently analysed some 0-60 runs on the data logger and decided to use the first two non zero points and then back extrapolated the data to the speed = 0 point to get a start time. Due to the difficulty measuring 0mph speed it seemed the best option for repeatable results. I then also interpolated the time for 60mph from the speed data. My results were:-

5.14s, 5.60s, 5.36s

I also did 0-100kph, reults were

5.43s, 5.90s, 5.66s.

These correlated with Racechrono times very well.

I'm trying to get some G force data for the shape of the start of the curve for my car, this will enable me to get a more realistic results as there is obviously a small transition between 0 and the clutch being fully engaged. I can use my method for comparison of data but I'd be interested to get more accurate times.

From looking at the data it seems that the start point is inbetween data points, thus anywhere in that 0.2s. My method assumes a sharp transition from stationary to full acceleration and this is not realistic. I think a realistic addition to the figures would be 0.1s to cover the unknown transition period, this will vary but over a few runs should avearge out.

aol_of_RaceChrono

Hello, RaceChrono tries to give more realistic start point, by some estimation and interpolation. I can't say how accurate the estimation is, but probably much more realistic than a start point from the zero speed.

Real G-force data probably would improve this estimation a lot, if you could make sure that GPS data doesn't lag behind. But how to do this GPS/G-force sync is an interesting research project :) Probably would have to try to sync the G-forces calculated from GPS with the real G-force.

Andy_Casey

Thanks for the mail.

I've got some GTech data on the way for the same type of car so will have a look what it thinks is hte transition from rest to full acceleration and get back with my results.

Andy_Casey

I've got the first piece of GTech data for a car of similar performance. From a quick analysis using 5Hz sample rate the error on the first 2 non-zero points is <0.1s and on the second 2 is 0.15-0.2s. This is for a 4wd car that does not spin off the line so it is really savage.

The intergration of the phone accelerometer to give the start point for non-zero acceleration would be a really valuable addition to the race chrono package as it would improve the performance testing suite for 0 to xx tests.

I've been playing for a week or so now and am really impressed with therepeatablility of the GPS data. I took this on as a gamble to save £200-£300 on a performance box as i like to number crunch and thus far it's had a lot of use! Got my £50 out of it already :-D

Andy_Casey

I've added the picture to the racechrono album I've created, p/w race_chrono:-

http://s348.photobucket.com/albums/q326/rusty_mcrusty/RaceChrono/

aol_of_RaceChrono

Very nice number crunching indeed! Thank you for your work! I will surely use your work when improving the performance testing section, will probably mail you some questions as well

Andy_Casey

Sure thing. I've just used first principles for data analysis. I need to look into more things to do with the power generation methods as unfortunatley the acceleration data is too noisy as it stands and is required assimply P = m.a.v .

I'm really pleased with the results and the correllation between racechrono and my methods and this means that I have confidence to use RC for comparison testing with the above adjustment for standing start times.

In fact I'll be comparing data tonight as the S4 tested before now has some modifications and we will be looking to quantify the gains. I will report back in a new thread and update the bhp thread if I can get some satisfactory methods...

Cheers, AndyC