How To Translate Steer, etc From EDR to vehicle responses

[MSI]

From INCR forum

• Q: I would like to be able to translate a steer angle as shown on the CDR report to the change in the heading angle from the original forward that would occur if the steering wheel were held in that position. Can someone help me understand how this might be done. I can use an exemplar vehicle to do whatever tests might be necessary, but if there is an analytical methodology I'd like to understand that first

A; First taking a look at straight forward approximations using simplified equations is a good start. As mentioned in the INCR forum thread and there are several others (i may post up some in a separate link next week, rattle my cage if interested)

I also recommend an excellent sophisticated analytical methodology which is to plug all the information from the EDR/CRD report into a simulation program and work to match each of the measurement points with the simulation program and then you have a much better understanding (and/or check on the simplified technique)

We have used this technique on a lot of EDR/CDR Report data for pre-impact and crash environment for live cases and for research.
Best cases are when we have information from BOTH vehicles...talk about additional pieces of the crash reconstruction puzzle.
Obviously my unbiased opinion (:-) is that the best simulation program out there is msmac3D

The main steps to use msmac3D to accomplish this task is...

1. Assemble vehicle information
   • loading? make/model/year to check source for vehicle specs (setup loads specs but always double check)
2. Assemble scene information as in slope/pitch variations of the terrain
   • Which can and will affect steer and braking responses
3. Digitize the output for the EDR data, first pre-impact(if available) and then the collision EDR (if available)
   • steer, yaw, accel lateral, longitudinal etc so can be read by a spreadsheet program (Excel, etc)
4. Set up the msmac3D(<-video How To link) simulation and input the steer and braking input tables to same as digitized data
5. Run the msmac3D simulation
6. Compare the simulation time history results with the EDR information
   • All variables in msmac3D are output in CSV format so you can easily open (with Excel, etc) and plot the data to compare with EDR/CDR
   • With Excel you simply link to the output file and visually compare the time history charts
7. Make adjustments to vehicle drag and/or steer to provide a better match
   • The CDR/EDR points are every 0.5, 0.1 or .01 sec increments
     • The simulation is computed every 0.001 sec (and output in any increment!)
   • Matching is a straightforward process
     • Iterate msmac3D to get as good a match as you can
   • Then for the vehicle crash information also do the same
   • Allow vehicles to impact and do the same on comparison of X,Y accelerations, Yaw pitch rates and angles (of available) x,y during collision
   • msmac3D calculates collision forces every millisecond,
     • This is in contrast to other programs with the simplifying assumption of instantaneous momentum exchange)

Once you get a match of the measured data you have an accurate representation of the pre (and collision) event and an ability to look at the whole picture (with simulation videos etc)
i will post up a sample next week (it's the weekend!)
But relatively straight forward and easy to do.
stay tuned!

[MSI]

The following is a quick run thru/comparison of msmac3D and EDR/CDR data.
In this case there were two events and we don't yet have the scene diagram/measurements/etc so more to come soon.

• Having scene measurements provides target marks to match, the roadway geometry and elevation changes (if any), curbs, etc

This is a quick look at the simple steps required to match steer, etc for the first event, more to follow...

To start we opened the Bosch CDR output for pre-crash Event 1 (Other data to be merged later):

2021-03-03_12-41-20.jpg (83.61 KiB) Viewed 1094 times

2021-03-03_12-40-42.jpg (240.97 KiB) Viewed 1094 times

To transform CDR data to useable comparison data:

• Print the CDR data as PDF
• From PDF editor save all information as plain Text
• Edit the saved text file
  • i like Textpad
  • Reduce the file to only include the various tables
  • I edited the titles to remove blanks so they'd appear in same columns as data
• Next import into EXCEL with blanks as delimiter (multiple blanks considered 1)
• In Excel separate the various chart data and create separate charts (see below for a sample of charts created)

Next we setup a msmac3D simulation with the vehicle year/make/model and loading
We then inserted the steer into the steer input tables per the CDR data above:

2021-03-03_13-21-51.jpg (33.22 KiB) Viewed 1094 times

inserted torque per the CDR report (above)

• The report provides timing and approximate values when you consider where measured/applied and total weight
• This was then adjusted to match the speed time history data:

2021-03-03_13-16-22.jpg (18.62 KiB) Viewed 1094 times

Then simply add the msmac3D output into each of the charts and compare
First we compared the speed, and adjusted the torque until we got an approximate match of the speeds

2021-03-03_13-11-01.jpg (21.1 KiB) Viewed 1094 times

Next we looked at the steer and yaw velocity

• no adjustments were required
• Steer was set to same as CDR report
  • Steering Ratio used since steer applied at the wheels, not steering wheel
    • If any issue with match of yaw velocity for given steer
      • check steering ratio, vehicle loading, moment of inertia, etc
      • we did not have to make any adjustments/variations

  

  2021-03-03_13-11-15.jpg (28.97 KiB) Viewed 1094 times

  Next we compared lateral and longitudinal acceleration
  • some issues you'll note below with the the longitudinal acceleration which is strange since the speed matches so well
  • Could be an issue with the sensor?
    Could the sensor location be the issue?
    msmac3D can monitor the acceleration AT ANY location so will investigate if that might be the issue

  

  2021-03-03_13-41-55.jpg (27.66 KiB) Viewed 1094 times

  Next is to merge in the collision information and then continue the comparison
  having a scene diagram with measurement tire marks provides a way to match the physical evidence
  More on this soon and we will add a video step by step too
  Here's a quick video of the vehicle movement...not too exciting...will add the crash, etc later
  Also hoping to find some better pre-crash and crash data/evidence for further demonstration of procedure and capabilities
  And then of course also what a msmac3D simulation can bring to help understanding and illustrating a more complete picture of a crash reconstruction.