The UnitsConverter cannot easily change physics and gravity volumes, as the binding to INI occurs at the class level and cannot be intercepted by a runtime conversion such as the UnitsConverter, given that we cannot arbitrarily subclass WorldInfo and override its PostBeginPlay to get our initialization code to run in its context. Given that the current gravity was determined experimentally anyways, scaling it should not bring too much grief if the only change occurs in the single INI file.
From: Balakirsky, Stephen B. [mailto:stephen.balakirsky@...]
Sent: Friday, July 22, 2011 10:29 AM
To: Brent, Taylor W.; usarsim-developers@...
Subject: Re: [Usarsim-developers] MassScale
For the second version, could we also use the UnitsConverter so that any future changes would be transparent to the developers?
From: Brent, Taylor W. [mailto:taylor.brent@...]
Sent: Friday, July 22, 2011 10:11 AM
Subject: [Usarsim-developers] MassScale
After much deliberating on how to deal with the mass scaling issues (different suggested values) we settled on using Sander's 0.2 value he got from gravity tests. A change was pushed to the GIT repository to implement this change. However, Stephen's experiments have shown that scaling with constraint forces differ from the force due to gravity. This results in two different scalings - 0.2 resulting from gravity tests, and 2.972 resulting from constraint tests. To fix this issue, there are two options we see: First, we can use the 0.2 gravity value as the MassScale, and change the UnitsConverter so that it scales all constraint values by 2.972/0.2. However, this would require a thorough find and replace/scaling of the MaxForce and Damping values we are currently using. The second option is to use the 2.972 constraint value for the MassScale and scale our GravityScale value (ini file) by 0.2/2.972. These two changes should ultimately result in the same outcome unless there is something we are missing. Can anyone confirm these changes would work as expected, and if so, is the second change suitable for everyone? Note the second change would mean going through any gravity/physics volumes and changing the values. Is anyone using gravity/physics volumes that would be opposed to this?