A recent post about crossing gates lifting mid train on another forum got me thinking about a few oddities I have noticed throughout the 11 years I have been playing some form of Train Simulator. 1) Crossing gates raising prematurely. 2) Cars driving through trains. 3) When starting from a stop locomotive speed increases, but the end of train is not yet moving. 4) Sometimes when stopping while the locomotive slows, speed sometimes the speed increases as the train slack runs in. So where is train speed monitored? 5) When using Third Rails Map sometimes it is the head of the train that is the focal point on the map and sometimes it is the end of the train. Does TSC really know where the end of train or how long it is?
I've always assumed that the loco moving off before the train end does is due to coupler slack. It seems realistic to me.
It might be the coach with the Driver Icon. - If you have changed ends on a MU you are effectively pushing the train
This is an answer from another forum. Re: Nagging Questions by kris120 ยป Tue Nov 21, 2023 5:24 am "All depends on the situation. 1.) Xings have their own scripts. These react only on the head and end of a consist (perhaps on DPUs in the middle, too, just do not remember). Most scripts close the gates when the head or the end is 100m away. Sufficient for neat little British trains, but not for something longer. Therefore when the head of the train has passed the gate the gates open. There is no check whether there are still wagons passing. 2.) Cars are only stopped by the scripts of the gates. Therefore when the gates open they pass the xings. 3/4.) The speed visible in the loco's cab, in F4/F3 monitor or F5 is checked only in the locomotive. There is a small difference between the speed in the cab's displays and F3/F4/F5 as the cab's displays are controlled by the loco's scripts and not by the internally coded TS itself. I don't know if the TS only considers the speed of the driver's locomotive or all wagons when scoring a scenario at the end. 5.) Don't know, have no experience. Regarding whether TS knows where the train starts and ends: Only as long the consist is not modified since start of the scenario." Running long trains on North American routes I have noticed in career mode that while running at track speed and applying heavy breaking that with the head end slowing down that the indicated speed will increase to the point that I get penalized points for over speed. This seems to be from run in with speed measured from the rear of train.
Might be the response given above, or the consist has multiple lengths. As it often works fine. Or the script is configured for typical train length for the route. Late raising isn't noticed. Crossing misconfiguration. If you play the All Aboard routes you'll notice that originally the game didn't have functional crossings. So he made a joke explanation about it in the manual: some drivers have excellent teleportation skills: you see them approaching the crossing, then miraculously appear on the other side (disappear, actually). Stationary cars are those without skills - often with people and dogs standing by, plus workmen. The old PDL has a crossing or two that may visually close (not sure) but cars keep crossing. I consider that one a feature, I keep honking at that silly impatient driver! Without actual knowledge I'm pretty sure that a section of track is connected to the crossing. The mistakes are similar to have signaling mistakes are there. Particularly on old routes with less building experience and way less powerful hardware. Slack and coupler elasticity (telescoping). Old releases have very loose coefficients. I've read that starting long trains would be impossible without this very real effect, as the trains, however little, do sit / sink in. Thus you want to compress the train before starting. Of course it's rarely an issue when distributed power is used, or the consist is relatively light. Only realistic to a small extent. It should never further under effort as the couplers are already stretched or compressed. Chain and buffers obviously do this as buffers are telescopes. See the pic below. Also needed due to curves. In unlucky cases you can self-collide and derail as the cars literally compress into each other. Using old trains with strong dynamic brakes adjusted for an entire lashup can end up like that. There is also an issue with the game where the consist randomly resets and other math hiccups which in absence of leeway cause consist separation. Hence, some releases break all day while others are entirely problem free. As I drive gently I don't find this an actual issue except when playing DLC with physics all over the place. The instant 5-10mph for some US trains is completely bull. You can easily observe it just watching the couplers and the distance between cars. FPS can also affect it as telescope math requires a few cycles: first the loco needs to gain speed to move and thus stretch the coupler which then applies resisting force which translates to acceleration which translates to speed which translates to increased distance meaning coupler stretch. Now, the first car is also moving, starting to stretch the second coupler. First part, the above, inverted. Second part: the loco that you drive. Helper service is a completely different experience - while the leading loco is having it fairly easy, the rear one does yoyo heavily. There are a few scenarios on Donner Pass and Horseshoe Curve to test it. Consist dynamics are the same but the monitoring is different. E.g. if you maintain 25mph on the lead loco while starting a 2.5% grade, the rear is certainly pushing and compressing up at 26-27 mph easily. During long mountain runs I often ponder the physics of energy, and how it transforms from one form to another: compression, momentum, heat. It's explaining why the train jumps as the rear loco arrives on a flat from an up-grade: it suddenly starts pushing, transfers compression and the wave eventually reaches the front loco, the whole train quickly gaining 0.5 mph.
