Do you have any msmac3D validation papers?

General Questions, answers and information related to the release of the msmac3D program.
MSI
Site Admin
Posts: 2303
Joined: Thu Jun 18, 2009 12:37 pm

Do you have any msmac3D validation papers?

Post by MSI »

Q: Do you have any msmac3D validation papers?

A: First and most importantly msmac3D is a combination of our extensively validated smac/msmac and hvosm/mhvosm programs. No conversion to other languages, this is the original code! with extensions and refinements.
  • Our many papers on those program include many demonstrations of validation.
    We are in the process of preparation of a rough draft of an 'official' msmac3D validation paper which we will make available on or shortly after the program is released.
    • (possibly as a white paper and then refined and submitted for peer reviewed publication)
    • Mainly because any and all correlation/validation simulations need to be rerun on the final ‘as-released’ program to insure no gremlins were introduced in the final preparation
      • For example, when working with another software company on a simulation project we found that nearly every update of the user input interface that they created produced different results! How? The user input interface they created arbitrarily garbled the inputs before passing them to the simulation program. So we found that different results occurred with nearly every update. And this went on for nearly 8 months!!
    • Now we don’t have that issue/problem at McHenry Software since we are doing the user interface and graphics programming ourselves and we do not garble inputs!
      However we want to be certain that any results to be reported/published are done with the final ‘as released’ version of the software.
  • We plan to create a ‘validation projects’ download section for our users (or might include as a project directory in the release version of our software) so users can test/rerun/etc. some or all of the validation simulations.
    • The main thing we are trying to do with this program is demonstrate correlation/validation with reality and present the complete input files required to produce the results.
    • We want to answer the important question for our users and critics:
      • How good does msmac3D correlate with measured reality and what did it take to produce the correlation?
      • Most validation papers out there do NOT include complete disclosure of inputs. Most 3D simulation validation papers out there include talk of correlation and some partial comparisons but none have published the
        • complete video comparisons,
        • the complete time history comparisons of roll, pitch, yaw, angles and velocities, accelerations, etc. and
        • most importantly the complete input files.
        Another issue:
        • name one program which has plotted the time history of roll/pitch/Z acceleration during a collision?
        • We don’t believe any have.
          • They mainly rerun the 2D simulations of crash tests and say ‘see…we are the same as the validated 2D’ or mention ‘good correlation’ but do not provide the complete inputs (one provided most of the specifications but no control inputs (steering and braking, etc.)
          In the rollover validations that they all only show up to the point of roll, and not how things correlate during the roll or rolls up to the point of rest.
      We are in the process of doing comparisons with all these metrics to inure our software has the best correlation with ‘reality’ that is possible.

      So sorry for the long winded response…
      • The short response is that we plan to publish correlation/validation papers with complete inputs and comparison of output metrics in short order after the release of the software.
        The video comparisons we’ve posted are preliminary demonstrations of how msmac3D using generic input approximations based simply on the msmac2D inputs correlate with reality.
        • We say ‘generic’ because for those posted correlation video comparisons the msmac3D program internally and automatically calculates all required 3D inputs based simply on the msmac2D inputs.
          And of course those msmac2D inputs are automatically calculated based selection of the vehicle type/year/model and adding any additional weight (an enhancement to the medit32 environment which will go out to all users when msmac3D goes out (meaning msmac2D users will also get the updates to the medit32 environment and graphics)
          Of course any and all of the 2D and 3D inputs can also be easily overwritten if measured values are known (like for spring stiffness, suspension clearances, etc.)
Be sure to also see Questions on the upcoming msmac3D Release

Thank you for your interest in McHenry Software.