ADR Data Acquisition – From Download to Meaningful Data We find a lot of people think accident data presentation is a one button process. It’s not Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Example – It’s not always what it looks like Personnel Process at the track Tools used Example – It’s not always what it looks like Data overlaid on video example Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Indy Car Accident Investigation Personnel One full time (Me) One permanent part time One rotating (per race) part time One consultant (Trammell) Jobs Performed at track Safety inspection on tech day (Helmets, Ear sensors, and Cars) Crash Box Download Pictures of crashed cars Database entry 2012 – 2015 (New Car) Average 70 Impacts investigated per Year Pool of 10 people to pick that rotating person from. Most are Engineers Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Tools Used for Analysis For graphing we use National Instruments - Diadem (Same program as Vandalia and CAPE use) For video photo analysis we currently use Adobe Premier Elements and Adobe Photoshop. (We often use VLC media player for playback) Pi Toolbox is used for chassis/engine data analysis Our crash box spits out Diadem files for direct import I’ve written custom script files for filtering and graph production Using diadem allows me to import files from our sled test into the same environment so I can do custom graphs from sled test for analysis New crash box will be excel based files, we will still use Diadem for our analysis Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Data Collection Process For Indy Car and Indy Lights only cars that come in on a wrecker are downloaded - To much damage to continue All cars are brought back to the garage area Once I see a car I can usually have crash data uploaded to our website within 5-10 minutes - Current system requires physical plug into car We also collect: - Videos from our Race Control system - Videos from the internet (if necessary) - Photos of crashed car - Track Photos (if necessary) Often make team shop visits to look at cars after they are torn down Consultation with Doctors (Trammell) Consultation with Safety Team (M Yates) Bone yard no longer used We don’t have the infrastructure to do successful wireless downloads Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Example Analysis Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Example of Rough Track Data Example from a rough street course race This is why airbags would struggle to work in one of our cars Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car .150 seconds Final output 1000 Hz Data Filtered with CFC 60 Filter Channel Frequency Class 3 db limit freq 100 Hz Stop damping -30 dB Sample freq > 600Hz We picked filter because that’s what the crash sleds were using .150 seconds Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Data Output from Crash Box Download Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car 60.7 G’s Lateral Example Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car Nose Input pulse was provided by nose of bullet car Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car Bullet car didn’t show similar G’s in the Longitudinal plane Need to look at resultants 25 G’s Longitudinal Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car Bullet car didn’t show similar G’s in the Longitudinal plane Contribution of Lateral Component ? Need to look at resultants 25 G’s Longitudinal Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car Even Bullet car resultants were not on the same order as the Target car All Peaks less than the 60.7 G’s Seen on Target car Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car vs Target Car Target car Lateral G’s Bullet car Longitudinal G’s (Inverted) Bullet car Longitudinal G’s vs Target car Lateral G’s When you overlay Target car Lateral G’s with Bullet car Longitudinal G’s the 60 G spike is unaccounted for. Even if you look Bullet car resultant G’s on the previous page they are still below 30 G’s at the same point as Target cars 60 G spike. Lower level of Bullet car G’s a result of only looking at single axis and not using the resultants ** Bullet car Longitudinal G’s are inverted for graph clarity Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car vs Target Car Crack in floor extends under crash box Crash Box location Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Speed From Video Target car’s speed just before impact with Bullet was calculated using distance on the track vs. time as calculated and observed from the video. The portion of the slide used covered 6 poles and 54 video frames. Fontana has not responded back on pole placement, it was observed from Google maps that they were 18 foot apart. Using this information Target car skidded 108 feet in 1.8 seconds for an average speed of 40.9 mph If 20 foot pole spacing is used then speed goes to 45.45 mph ** The next slide is screen shots from video used for above calculations. Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car vs Target Car 1 2 3 4 The struggle in calculations like this is with the slow video rates Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Speed Calculation Speed when front wheel starts spinning again as observed from video, is when Target car and Bullet car separates Speed at this point is 101 mph Delta V, 141-101 = 40 mph over time of contact Because of the nature of this accident we were able to validate the Delta V calculation. The Pi data shows us the beginning of contact and the separation of the two cars The math show a 40 mph delta V Bullet Negative Longitudinal G spike Indicates hit with Target car Speed at this point 142 mph Bullet Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Bullet Car Delta V From Crash Box Calculated Delta V approx. 38 mph Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
Target Car Delta V From Crash Box 3 axis approx. 56 mph 2 axis (X,Y) approx. 51 mph Higher Delta V for Target car is probably a result of Bullet car “punting” it forward from impact Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR
James Jakes Data Over Video Privileged and Confidential Do Not Copy or Distribute Without Express Written Permission From INDYCAR