xms calculation

[shefalika]

I am trying to calculate Chest Clip for a dummy. I use the Xms calculation to calculate that over 3ms but for some reason Diadem does it over more than 3ms. Is there something that can be done to change that?

[IS]

To be able to answer I need some more information: First of all, which version of DIAdem are you working with and then, what is the resolution of your time-channel?

[shefalika]

I am using Diadem 8.0. As far as the resolution of the time channel goes, it is in seconds, 5500 points, starting at -0.05 and ending at 0.0025, which point differnce being 0.0001. I hope this answers your questions. Please let me know if there is any more information needed.

Thanks a lot.

[IS]

OK, I see. The resolution of the time channel should not be the problem. I actually could reproduce that myself testing with the file crash.dat in the DIAdem library folder. The second of the crash channels produces a 3ms of 3.07 sec.
But if you have a closer look at the online help you will find out that this is perfectly allright. According to the description of the implemented algorithm (NHTSA_FORTRAN-Code), the Xms "computes the maximum value that the linear interpolation of a time series,..., *meets or exceeds* for an interval of at least 3 ms,..."
Thus, if you calculate a 3ms value the length of the resulting interval will be at least 2ms, but might be longer as well.
This is how the Xms calculation is defined within Crash analysis. If you need th
e function to behave differently I am afraid you have to figure out another way of calculating.

[DIAdemon]

The XMS caculation will always return a value greater than or equal to the target duration.

For the "over multiple peaks" option, the time values (interval start and interval end) are actually irrelevant. XMS simply sorts the channel with respect to amplitude. It then saves the highest N points (enough points to satisify the target duration) and disregards the other points. The amplitude returned is the lowest amplitude of the remaining points. The time interval returned is the earliest and latest points of the group. This represents the N points with the greatest duration. The points are not necessarly continuous and are smattered across the graph. That's why I do not try to construe any information from the time, only the amplitude is relevant.