Product Support

Who is Bill Mitchell?

Data Acquisition Procedures

The following are suggested data acquisition procedures based upon a decade of doing analysis in the field (or in the front of transporters):

1.  When to do data acquisition.  Do not bother with a data system until your team has mastered the art of preparing the car, getting it to the track, and putting a reliable car on the grid.  Your first priority is preparing the car, not collecting data.  You should never omit basic preparation tasks such as alignment because you are working on the data system.  If the data system detracts from the basic tasks, either get more crew or take the data system off the car.

2.  Testing the system.  The data system should be tested before you leave for the track.  Routine maintenance, engine changes, or tire changes may have dislodged the sensors or connectors.  Turn the system on and record data while you move the throttle, turn the steering wheel, shake the accelerometer box (which contains the G sensors) and jump up and down on both ends of the car.  All of this motion should appear in the data you collect and examine.

A simple way to calibrate the accelerometers (G sensors) is to rotate the box 90 degrees and see if the appropriate G values read +1.0g or –1.0g.  They should also read 0.0 when the box is installed and the car is level.

Engine RPM should be tested when the engine is tested or warmed.  It is very easy to move a mechanical RPM sensor during engine changes.  Test the system when the crew starts the engine.

Wheel RPM is also easy to dislodge and hard to test.  It is very difficult to physically spin a tire fast enough to record a wheel RPM value.  One alternative is to place a magnet on an electrical drill and spin the magnet past the sensor.

I also like to test the system each morning before the car goes on the track.  This can be a useful test, plus if something fails you can honestly state you tested the system before the car went on the track.  I do not usually test the system between sessions.

You should also be personally responsible for all the computers, cables, PCMCIA cards, printers, printer paper, ink cartridges, blank disks, and tools needed for the data acquisition system.  Have spares for as much as possible.  Do not let the data system be a burden on the rest of the crew.

3.  When to collect data.  You should collect data EVERY time the car goes on the track.  This includes warmup laps and every test session.  You can never tell when the car will not return from the most basic trip and you want to have data.  If nothing else the warmup laps serve as a test of the data system.

4.  First look at the data.  For the first look at the data I like to look at the entire run for important events like engine over-revs.  These can appear as easily on a slow lap as a fast lap and can easily occur on out laps or in laps.  Over-revs should be reported to the crew chief as soon as possible.  You should report how high the engine revved, how long it stayed there, and the circumstances.  Was the over-rev on a downshift or a missed up-shift?  How long did the over-rev last and was the driver on the throttle during the over-rev?

Take a quick look at all of the sensors and verify they all worked throughout the entire session.  Also look at channels such as water temperature and oil pressure.  A flat line usually means the sensor was not working.  Any abnormal values must be reported to the crew chief.

Look at maximum speeds and check the engine RPM at the fastest part of the track.  The maximum speed may not have occurred on the best lap.  This is your first indication of whether you have the proper gearing.

5.  Look at the fastest laps.  Now load the fastest two or three laps as well as any laps from another session.  The first variable to examine is speed.  See where you have gained or lost speed compared to your baseline lap.  The second variable is usually engine RPM to verify gearing and engine performance.

6.  Verify conclusions with other laps.  Avoid basing conclusions on the fastest lap alone.  Look at other quick laps to verify what you see in the fastest lap.

7.  The Driver Debriefing.  I like to look at the data during the driver debriefing.  If the driver mentions a gearing problem or handling defect you can often look at the data and support or explain the driver’s comments during the debriefing.

8.  The Race.  Surprisingly little time is spent looking at data from the race.  After the race the emphasis will be on loading the race car and preparing or the next race.  It is often impossible to look at data right after the race.  I like to perform a quick analysis right after the race to look for over-revs or other significant events.  Additional analysis should be performed back at the shop.  It helps to have a race log which describes where the car was on each lap (running alone, in heavy traffic, lapping slower traffic).  This log can explain the data.

9.  Pick up the timing beacon.  One of the most common mistakes is to leave the timing beacon.  In the rush to load the transporter it is easy to forget the timing beacon at the edge of the track.  You can avoid this by tying a large red flag to a portion of the truck, just as aircraft use large red flags to denote a plane which is not ready to fly.

10.  Write a report.  You should always write a data report similar to the engineering report prepared by the team engineer.  The data report will help you consolidate what you see during the weekend and will be a great benefit when you return to this track.  A five to fifteen page report will assist you greatly when you return to the track.  The factors that are so evident during the event will be forgotten before you return to the track.  With a written report you can refresh your memory in ten minutes.

I like to capture and print every screen which is closely examined at the track.  With Windows this can be done easily IF YOU GET IN THE HABIT.  The report can be printed at the track or after you get back to the shop.  I send reports to the owner, team engineer, crew chief and driver and keep a copy for myself.

11.  Capturing Screens.  With Windows it is fairly easy to capture screens and paste them into a report.  When I have a screen I want to keep, I hit Alt-PrtSc to paste the entire screen into the clipboard.  I use the Notebook feature to add comments and cut the screen down to the portion I actually want.  I use the Start Box command and the mouse with left-button to draw the portion I want to keep.  Then the Ctrl-C commands cuts the text and places it in the clipboard.

I have a copy of Microsoft Word open and I paste (Ctrl-V) the clipboard into the open Word document and add descriptive comments, such as "over-rev during morning practice.”  Be sure to add enough comments that the meaning of the graph will be evident a year from now.  If you get in the habit of doing this your report will be nearly written when you get back to the shop.  If you don’t do this you can easily spend a full day writing a report.

12.  Experience Counts.  Extracting information from data is largely a matter of experience.  It helps to have knowledge of driving techniques and the basics of vehicle dynamics, but the more data you look at the more you will see in the data.  This manual and Buddy Fey’s book Data Power can accelerate your learning curve, but experience is the real factor.

You should develop your own task list of things to look for in the data.  You may start with Speed and Engine RPM but develop your own list of things to look for every time the car runs.  This list may vary with the data system, the car (formula, sedan or showroom stock) and the track (road course or oval).  Always go through your basic list.

After doing your basic list, spend some time “mining” the data.  Look through the data to see what else you can learn.  This may be a reaction to problems the crew is chasing or the driver has reported.  It may be a mysterious miss or a handling problem.  If there is no particular problem to chase, then see what else you can discover in the data.  This can often be very fruitful for this event of future events.  And if nothing else if allows you to keep busy.  Never look as if you have nothing to do.

Tech Reports