Physics

This page needs to be formatted into a decent information source, with commentary on what is simulated, what is not simulated, and what the suspected bugs are if any. -chris

Coupler Breakage

The coupler breakage feature was a great idea in Trainz Railroad Simulator 2006 for the realist but it has problems with its script that doesn't allow a different calculation for stretch,expansion and buff,compression forces. Trying to include the "coupler-compression" in the script would do nothing. A train just moving down a track should not break a coupler aka "knuckle". The buff,compression force on a standard "Janney" knuckle is huge and the car is likely to be crushed or jackknifed before the coupler breaks. Under normal freight train conditions the slack is stretched, like a chain, through the whole train and when the train tops a hill and the last car(s) start accelerating and it can actually get going faster than the locomotive but only until the slack is bunched all the way to the loco, when the slack hits the loco it accelerates to the trains speed and then the slack is bunched. This happens not all at once but car by car starting with the last one and acts like dominoes. The tensile strength of a knuckle is around 390000 lbs ~1735kN static shock load can reach 520000 lbs or more.

I'm uncertain on what your point is here. Trainz calculates the compression and expansion forces running through the couplers, and flags them as broken if these forces exceed the specified limits. You can tweak the limits to whatever you think is appropriate. Slack and compression of the entire consist is modelled appropriately to achieve this result. -chris

chris, the point is couplers don't break in compression, cars are crushed or most likely jackknife aka derail. I know from experience that on a 13000 ton coal train, that when the train brakes are used even in an emergency application that the couplers don't break. Also trying to use dynamic brakes on the engine also break couplers and it usually happens between the engines, if the engine isn't powerful enough to break couplers in expansion then it should not be able to do so while dynamic braking. I was going to find a few pictures of where couplers have hit head on at high speed and the coupler survived the car or engine was crushed. i just couldn't find one on the net. when i try to use the physics, no matter how high i make the settings the couplers will break in compression the stock settings are 2500kn in expansion; there are no compression settings without changing the script manually; and I've tried 100000kn, and thats not a typo, and still break couplers in compression. The problem is, is that trainz does not separate compression from expansion calculations. if one or the other reaches the limit, the coupler breaks. Vehicle Physics,<kuid:-16:280742> is my version that is doing this. if you need more detail ill be back. thanks chris.

From what you're saying here, it sounds like we're modelling both compression and expansion correctly (and yes, we model these seperately) but we're not modelling any kind of compression within the vehicle itself. This is quite true - we assume that the vehicle is infinitely rigid, so any movement applied to one end of a vehicle will result in an equal movement at the other end of the vehicle. This will certainly exaggerate the forces on the couplers, even if the coupler calculations are correct. I doubt that there's a fault in the actual coupler calculations, because that's just a measurement of the force as a byproduct of the movement simulation, rather than a seperate simulation for the purpose of coupler breakage. Remember that we do not simulate any destruction of the vehicles, nor do we simulate any off-center forces on the train other than the effect of curved track, so if something is going to give internally, it will be the couplers. -chris

ok Chris here it is, a picture of the head on where the coupler survived and the engine was crushed. Head on btw I'm not an English major and don't know how to format, but I know physics and trains.

Please Chris - if we're going to have a Wiki at least make an effort to actually read what is written. The previous contributor just stated that even when he sets the breakage value under compression to a ridiculously high figure, the couplers still break. You then say it sounds as if everything is working correctly?!

Please re-read what i said above. -chris

The other problem is that despite what you say Chris, the current implementation does not break couplers if the specified force is exceeded in tension. The simple test is as follows. On the Tidewater route there is a grade that climbs from the river up to the tunnel (the one you guys used to do the steam physics demo session in TRS2006). On this grade put an SD40-2 and the 3000 tonne US Coal consist. Add the vehicle physics rule to the session and set the coupler breakage value in tension as 1Kn. The standard SD40-2 enginespec throttle curve puts out 165Kn tractive effort at zero speed in notch 1, so as soon as we release the brakes and open the throttle to notch 1, we should have exceeded the first coupler's breaking tension by a factor of 165 times!! They don't break though.

I personally suspect that Trainz actually calculates neither tension or compression forces correctly. My test could likewise be used to test compression sinply by putting the loco at the back of the train and pushing. The only time I've had couplers break is when there is a speed difference between two cars equal or greater to the speed value set for "couplers break when coupling above X mph" in the rule. -Stuart

The forces at the couplers are not a separate calculation, but rather are an integral component of the train movement physics. If this was not functioning, the train would not move. So I think it's reasonably safe to say that the actual forces are being calculated correctly. Now, that doesn't mean that the coupler breakage code is working, or that the rules which control the coupler breakage setting is working - so I am not saying that there are no bugs here, but rather that I don't think that there are any problems with the coupler force calculations themselves. If you have a reproducible case which demonstrates a failure of one of these components, then I'd strongly encourage you to bring this up on the public test forum where the QA guys will see your comments and follow it up. The wiki is a documentation tool, not a bug reporting tool. -chris

Cab Sway

The cab sway was also another idea that never worked in TRS2006 no matter how much you changed the settings and should be addressed or removed. -dmatlock

It works as designed. If you'd like to see a nicer effect, please specify details about what you think the effect should be. -chris

chris i can't tell that the engine(in cab mode) sways or cars for that matter they all hit a turnout and its like nothing happened. in real life I've never seen a car or engine not sway even when going on straight track. all cars/engines sway and most have a resonance between 22km/h and 35km/h and can sway enough to derail under the right conditions. the taller the car (you guessed it) the more it sways. speed, cg, and springs on the trucks all have a little to do with it, even the wind and direction have an effect. if trainz really wants to make this part work they must have the cg of the cars and engines and the cg if its loaded and calculate the sway of each. Man thats a lot of calculations just for physics, and thats not all thats needed to make a train in trainz look, act, and feel like a real train. later chris. -dmatlock

I feel I should mention that I have seen cars and engines with the sway working but the sway has been way to much; more like a ship in a storm than a train going across the desert. -peter

ok chris ive looked at the "sway" again and it does have an effect but its more leaning than sway. when i think sway its not steady, but varies as the springs compress and release. -dmatlock

I think that sway must be simple, not to overload the core of the game. It may be created to join tracks or a junction. In such a place would train made sway, which can be defined in CONFIG. This definition is only a "solid spring", nothing more. Simply setting from 0 to 1 is needed ;-) Big sway may also be made manually, using a set connection tracks - its downgrade/reduction (I make it in TRS04: not bad results :-) Professional term is (cross) Fall. Sorry for my school english :-) -drew