after a lot of testing and research, I've come to some conclusions about resistance in RailWorks (apart from it being implemented in a
way, that is), and how one can more-or-less "cheat" his way to prototypical values.Before I come to the actual point of this topic, I think it would be best to post a link here which could prove very useful to modellers:
http://babel.hathitrust.org/cgi/pt?id=m ... 1up;seq=44
Starting from here, there are a couple of pages from a Baldwin book which should give everyone (even me) a good understanding of the various resistance losses occuring in a train, together with a very useful, and probably rather accurate formula to calculate the resistance of a train.
Now to RailWorks. Friction in RailWorks is implemented via 3 constant parameters and one variable.
These are:
1)mass
2)rolling friction coefficient
3)drag coefficient
4)speed
Although one would expect mass to be a rather straightforward thing, it'd probably be a good idea to mention here that the mass in the configuration files is in imperial (long) tons.
The rolling friction coefficient makes it possible to create a constant retarding force which acts upon the train. A value of 0.01 here would mean that you need to use 1 lbf of force to move 100lbs of weight at a constant speed over the rails.
As we know that the Davis equation mentioned in the link above is a 2nd order/quadratic equation, we can substitute the long mathematical thingy above with the formula R=a*x^2+b*x+c. The force caused by the rolling friction coefficient thus takes the role of "c".
The drag coefficient should thus take the role of "a"... But wait, where's "b"?! Well, gentleman, the short answer is RailWorks. Which leads me to the main point of this topic:
Although the drag coefficient can be used to simulate both the "b" and "a" components of the Davis equation, the results are often somewhat... off.
To demonstrate this, here's a chart I've made with the help of Excel, trying to find a neat way of circumventing the limitations of RW:
The greatest absolute discrepancy is around 49lbf at 35mph, the greatest relative discrepancy is however a terrifying 24,515% at 20mph.
Here's where the community kicks in! If people are willing to take their time and modify the above named parameters to achieve more accurate results, they absolutely need a reference frame!
As you see here, dear forum community, as the 2nd defining point of this chart's equation and drag coefficient, I've used 70mph. I think we will all agree that WD 2-8-0 Austerities haven't ever reached 70mph, although this is possible, with a very light load or favourable gradients.
That's why I'm asking you guys to post here highest average running speeds of specific equipment, at a specific time period, and with a specific operator.
(In case this doesn't sound to clear to you, I mean something on the lines of: Milwaukee F7 4-6-4, 100mph*)
*100mph on schedule, yes they reached more, so if the community is willing to, we could create some sort of standard to bring maximum accuracy to the operating speed range of the equipment we have available. Whilst it won't entirely remove the inaccuracies forced upon us with RWs physics model, it should atleast bring us closer to reality.
Thanks for your time, and sorry for my poor English.
over accuracy like me, your best bet is to take a drawing of the loco/the frontal part, determine how many pixels the distance between the 2 wheels is (4ft 8,5in), and then patiently cutting out little boxes from the drawing in paint.net, calculating their area as you go)*V^2
I think I'll stick with the book's default 120ft^2 for the effective cross-sectional area. Davis made it a standard for some reason and it'll make calculations a little less time-consuming.
