Re: [FT] Scales (was: RE: [FT] back to fighters)

Roger Burton West wrote:

> On Tue, May 14, 2002 at 03:16:01PM +1000, Robertson, Brendan wrote:
>
> >Not bothering to work out if the math is right:
> >5 minute turn; 1 mu = 200 km; 1 thrust = 1g ?
>
> Not the same thing as Oerjan's suggestion. The basic problem with any
> scale system is this:
>
> If I accelerate at 10 m/s for 10 seconds, I'll be going at 100 m/s.
> However, I will have covered 500 m.
>
> If an MU were 1000 m, a turn 10 seconds and a thrust point 10 m/s,
then I
> ought to move half a MU on the turn I accelerate, then 1MU each turn
> after that. If a thrust point is 20 m/s, then I move 1MU on the turn I
> accelerate but should be going at speed 2 after that.
>
> Therefore, whatever numbers are chosen, you won't get a physically
> accurate result. It's probably a necessary compromise. (In essence,
you
> model the acceleration happening all at once at the start of the turn,
> rather than spread throughout the turn.)
>
> thrust = (MU size) / (turn length ^2), if you're using the velocity
measure
>
> Double that if you're using the distance measure.
>
> Roger

This just goes to show how amazing the inertial compensators are. 
Although the
average acceleration is a measly 1g, the acceleration is actually
delivered in
almost instantaneous bursts of almost infinite impulse (thrust is is
delivered
following a Chronecker-delta function [infinite magnitude, zero
duration,
finite integral]), so the velocity changes in an almost steplike manner.
Fortunately, the inertial compensation is actually a side effect of the
main
drive operation, and there is no need to actively balance their
operation.
Poor design of the impulse envelope produces results that vary from
comic to
tragic.  Errors of not extending the envelope back far enough are self
correcting, as trailing portions are simply left behind, but
unaccelerated
forward portions smash through the vessel, much like a retributive Fist
of God,
or at least a finger poke.

[Can you imagine trying to play with constant acceleration?  How many
templates