Videos that Merge Real Footage and Simulation/Animation

[MSI]

A recent video posted to the internet which showed a

• vehicle flipping in response to an interaction with a loose tire
  • we captured and reposted with some slow motion, etc
  • We determined that the loose tire interaction was FAKE!
  • The technique used was to merge live video footage with animation footage.
  • Looks great...but...NOT real

That video reminded me of a case we did in 2005 (depicted in the video above):

• Note the above video with excerpts from a 'crash reconstruction' video (the real looking video) an expert attempted to use as evidence of their crash reconstruction.
  • Also are frames from some of our analysis of their pc-crash simulation.
  In the case, once we saw the video we asked for documentation of their crash reconstruction which supported the movements of the vehicles.
  
  • They had used pc-crash for their crash reconstruction analysis
  • We requested they document their pc-crash results, see WHAT YOU NEED to Evaluate an 3-D Reconstruction/Animation
  • We received the inputs and outputs of their pc-crash analysis and so we reran it to verify that the results were as they displayed.
• To further test their results we also ran a msmac simulation.
  • With msmac the crash forces and moments are considered for every millisecond of collision interaction.
  • With the pc-crash program the collision forces are reduced to an instantaneous exchange of momentum
    • most vehicle crashes actually take from 75-200 milliseconds
  • When we reran their crash reconstruction with same speeds, weights, offsets we found that the vehicles did not behave as depicted in their animation or their pc-crash results.
    • And as a further part of our investigation we found that the automatic secondary collision option of pc-crash ignored ALL the secondary interactions between the front of the truck and the rear of the limousine.
      • this case also had scans of the vehicles and so when you viewed the movement of the vehicles subsequent to the initial impact you found that the structures of each vehicle were interacting which was not caught with the secondary contact option of pc-crash.
  When those forces and moments of the secondary contacts were properly addressed the speeds and offset had to be changed to get a match of the vehicles at positions and orientations to match the evidence.
  This essentially invalidated the pc-crash reconstruction basis for the movie (see above) and the video was not allowed to be shown in court

we may add some additional information on our analysis at a later date.

MAIN POINT: be careful using the automatic secondary contact option of pc-crash.