I've started to work on a conversion of pak128 to experimental. In this thread i would like to outline a few of the concepts i use for the balancing. In a pak like pak128 with vehicles from all over the world, the main aim is not to have real world economic values, but internal consistency of the pak. the project is at github: https://github.com/sdog/pak128-experimentalload times for pax
The plan is to calculate load times based on the number of doors on either side n
and passenger capacity p
. The width of the door should have a non-linear relationship to the throughput. This will be neglected though. Thus a single door at an old long range trains has a value of 1, double doors 2, i shall give wide double doors (eg in subways) a value of 3.t = c n p/3 c
is a constant conversion factor, to be tweaked to get results compatible to the game.running cost
per km running cost of vehicles is to be calculated based on the engine power, combustion type and (perhaps fuel type). The primary energy required for one km at 2/3 maximum engine power is calculated at half maximum speed v
. The monetary cost is a time dependent function, normalized to 1000 for year 2000. A constant factor c
that is tweakable is used to convert to simutrans money.
Unpowered vehicles need more thought. Two approaches might be sensible. Guess acceleration profiles and base cost on kinetic energy (at half max speed) or on friction only, introducing friction/drag coefficients.
Beware that all of this is super crude, averaging over non linear relations is a bad
thing to do.monthly vehicle cost
regarding monthly vehicle cost, please find first concepts in this thread:http://forum.simutrans.com/index.php?topic=7937.0
At this point in time i collect data on powered rail vehicles. As soon as i have entered enough to have a good variation for testing, i can start to write the scripts for dat creation.to be continued in later posts to this thread