Are long consists impossible?

[MikeK]

Thank you Kali, I think this validates the theory.

With your modified coupler I was able to use as much dynamic brake as I needed to hold train speed with no derailments. The DBs are still way too effective, but that is another discussion.

There is no more slack in the train but that is obviously due to the improper modelling of the couplers as springs with no limit to their movement.

Well, almost no slack. as you pointed out, the loco couplers still have original settings:

Full DB:

2011-11-21_00005.jpg

Followed by full throttle:

2011-11-21_00006.jpg

So it looks like we can have "slack" in the train (although not realistically modelled) and we have to steer clear of dynamic brakes, or we can remove the slack and use all the systems on the train.

Frankly, until RS.com can fix the way couplers are modelled I think I can live without the slack. I would rather have things missing from the sim than have them incorrectly modelled.

With these couplers, weaker dynamic brakes and less effective train brakes, it might be possible to have a realistic driving experience after all.

Is there any easy way to apply these same settings to all the loco couplers as well?

I wonder if this is also related to the derailment issue that b737lvr originally had when building a long consist ...

[Kali]

I think all the US stock uses a copy of the default coupler with different paths in it, so editing theirs should bring no surprises. If you can do that you can also open up the simulation files for the DLC engines and cut their dynamic brake values to something useful, like 50 instead of 400. If you try the default SD40 or the ES44 now you'll notice it uses the "fixed" coupler. Does anyone have any real drawbar pull numbers, btw? I stuck the spring coefficient I used in that blueprint in the spring equation, and it seems that if I accelerate my ( 8000 ton ) coal train at 0.95m/s^2, which is about 0.1g and a bit much for a coal train I'd imagine, it will stretch 10cm. So it probably still isn't quite stiff enough. ( a 12s 1/4 mile dash is a bit over half a g on average for comparison ).

I think I settled on 45% for brake force value for the coal cars, so that's probably ok for any other car that you can load - for one with a load built into it's mass you want more like 13% or a bit less. Brake force appears to only be a percentage of the given mass, not the mass+load.

Why I mentioned coupling heights is because I was watching the head car when I was doing some brake tests, and even though the second one has very nearly the same force acting on it it wasn't looking like it was about to tip over like the head one. It was hard to see - and then I had to give up with a headache - but it did look like the engine it was coupled to had it's coupler slightly higher, which is going to do odd things to force vectors.

[MikeK]

Update: I applied your modified coupler settings to the loco that I was using.

By playing around with forward then reverse then forward then reverse and observing the head end speed reading while watching the rear of the train, I estimate that there is about 2 car lengths of slack in my 100 car train. Probably as good as it is going to get I think.

Thanks again Kali! I can play this game without the risk of my computer ending up in my neighbour's yard! :)

I wonder if the RW_Tools author would be able to modify the tool to apply a standard more realistic set of coupler settings, freight car brake settings and dynamic brake settings to all locos and cars in one go? I am sure that steam will override the current settings at some point.

[Kali]

The problem with the brake settings is it's dependent on the wagons being loadable; 45% is the conclusion I reached for a loaded wagon given that it is set up empty. 45% for a wagon that is set up pre-loaded is 300% of what you want it to be.

We can probably put a coupler pack together you can just reinstall after a verify, there aren't that many coupler blueprints - first though let's get it as good as possible. Getting RWTools or another program to standardise coupling height might be worthwhile.

[MikeK]

After playing around with the settings some more, I have ended with these values: 20000 for the spring coefficient, 10000 for the damping.

With these value, I can still see couplers visually separate a tiny amount if I go to max forward throttle while the train is moving rearwards (extreme case), and on a long heavy train I can still derail a train If I run the slack in too quickly. But as long as I "ease" the slack in I can end up at 100% dynamic brakes, which is what I think is realistic.

I think these settings are a good compromise because with a 100 car train it gives several car lengths of slack that can run in and out, but the higher damping stops the retarded slinky back and forth effect.

When transitioning from power to braking you have to use a low dynamic brake setting for a period to run the slack in before upping the brake settings, otherwise you risk overpowering the damping and causing the derail. Similarly, when you transition from dynamic braking to throttle you can watch the speedometer and see that the slack is clearly running out before the head end speed will start rising.

Overall I think these settings are a good compromise, considering real couplers are not unbounded springs, so it will never be truly realistic without some work by RS.com.

But with these settings I get what I really want: to be forced to think like a real engineer, using gentle power and brake transitions to manage slack and always be aware of whether the train is stretched or bunched, and to take action to stretch or bunch it to handle upcoming grade transitions.

Honestly I don't care how unrealistic a sim is if it can force me to think realistic thoughts while using it. If the unrealistic elements get in the way of that then it must be fixed!

[GaryG]

When transitioning from power to braking you have to use a low dynamic brake setting for a period to run the slack in before upping the brake settings, otherwise you risk overpowering the damping and causing the derail. Similarly, when you transition from dynamic braking to throttle you can watch the speedometer and see that the slack is clearly running out before the head end speed will start rising.

Hi Mike

That method is exactly the method recommended in the real world. Power off, wait ten seconds, put DB in Setup and once engaged bring up the DB setting a bit to bunch slack and then you can apply the setting wanted. I have read that the ammeter is watched for the bunching slack setting; I don't think we can use that method (yet).

It certainly does sound like excellent progress is being made with the testing you fellows are doing.

GaryG

[krellnut]

Thanks a ton for this,guys. I want to showcase my wip SD70 sound pack, but with that stupid coupler stretching thing going on, I was going to show it in Railworks2. But I just tried your coupler numbers and now the stretch is gone, awesome work.

[Griphos]

Tinkering in files is a mystery to me, but this is very promising, so I'll do some backing up and experiment with these modifications. Thanks for all the really smart discussion!

[MikeK]

What would be really cool is if RS.com could do this properly and add "MaxStretch" and "MinStretch" parameters to limit how far in and out the couplers can move before they stop. Then we could turn the damping and spring rate back down and have cars that really did bounce back and forth like they do in reality.

Also, they could play the "coupler collision" sound whenever a coupler reached the max or min value. Then we would be able to do something like this:

http://www.youtube.com/watch?v=SaF82LCBSyA

Listen at 6:40 and 16:58

You could use the 8 view to move halfway down the train, add some throttle and hear the couplers stretching starting from the front of the train and moving towards the rear.

[Kali]

Well, the spring equation - Hooke's law if you want to look it up ( it's a good bet they use that, most of the other physics seem to use basic real-life laws ) - shows if your coefficient is right at some point the stretching force needed to stretch any further will be more than the engines can generate, which is both your max and min stretch. If you sit down and work the acceleration of your consist out - which is merely measuring how long it takes to get to a certain speed, you don't need to be *that* accurate here - you know how heavy it is already so you can work out the force needed, and you already know how far you want to allow it to stretch, you can work out a likely number for the spring coefficient.

[MontanaRails]

Yeah, there's been funny couplers for quite some time - ever since people asked for coupling slack, I think. Probably about time someone experimented with them a bit. I would think the spring force needs a couple of 0s on the end at least, perhaps when the cars compress the couplers under braking the collision boxes are overlapping.

The default SD40s dyna brakes aren't terrible - you might want to try the ones I posted somewhere in the "dodgy physics" thread which have been reduced a bit again. Note I posted two lots, try the second one. And yes the car brakes are too strong, we discussed that there too.

Edit: default couplers live here:

Assets\Kuju\RailSimulatorUS\RailVehicles\Couplings\Buckeye\Type-E\buckeye_type_e_coupling.bin

and I'd imagine the lines to mess with would be these:

<SpringCoefficient d:type="sFloat32" d:alt_encoding="0000000000708740" d:precision="string">750.0000</SpringCoefficient>
<Damping d:type="sFloat32" d:alt_encoding="0000000000C08240" d:precision="string">600.0000</Damping>

They did change things, then, with TS2012. Heres what the file looks like from RW2 (when we didnt have the stretchy coupler issue - but did have slack action). Just an idea...I think if we leave the min distance and max distance both set to "0" it acts as a rigid bar (like a tender/steam loco connection). If we leave the max distance set to .1, and set the spring coefficient to 0, does it then use the hard limit of .1 and ignore the spring? I'm not able to test it right now...

Code: Select all

<?xml version="1.0" encoding="utf-8"?>
<cBlueprintLoader xmlns:d="http://www.kuju.com/TnT/2003/Delta" d:version="1.0">
   <Blueprint>
      <cCouplingTypeBlueprint>
         <UncoupledGeometry d:type="cDeltaString">Kuju\RailSimulatorUS\RailVehicles\Couplings\Buckeye\Type-E\[00]buckeye_type-e_uncoupled</UncoupledGeometry>
         <Bogey d:type="cDeltaString"></Bogey>
         <Strength d:type="sFloat32" d:alt_encoding="0000000000C06240" d:precision="string">150.0000</Strength>
         <CouplingConnection>
            <cCouplingConnectionBlueprint d:id="27636024">
               <Type d:type="cDeltaString">buckeye</Type>
               <CoupledGeometry d:type="cDeltaString">Kuju\RailSimulatorUS\RailVehicles\Couplings\Buckeye\Type-E\[00]buckeye_type-e_coupled</CoupledGeometry>
               <ReceivingGeometry d:type="cDeltaString"></ReceivingGeometry>
               <PivotType d:type="cDeltaString">eMidPoint</PivotType>
               <Damping d:type="sFloat32" d:alt_encoding="0000000000000000" d:precision="string">0.0000</Damping>
               <ManualCoupling d:type="bool">0</ManualCoupling>
            </cCouplingConnectionBlueprint>
         </CouplingConnection>
      </cCouplingTypeBlueprint>
   </Blueprint>
</cBlueprintLoader>

[MikeK]

MontanaRails, did you find anything with those max and min values? I played with them and saw no effect. I am guessing that they might be broken at this point.

[Kali]

Are you sure those aren't something to do with actually coupling?

[Griphos]

I just took a longish consist up from SanB to Summit. I had 2 ES44AC and 2 SD40-2 on the front of 71 stacks, mostly double. I use Britkits replacement buckeye coupler. I haven't gone down hill yet, but the trip up was fine. No unrealistic stretch, but then I never was above 500 amps, even at 100% throttle on the 3% grade near the top. I was able to stay near 30 mph all the way, so I wasn't struggling. Train was tight all along its length. I'll add another 30 or so cars when I get the chance and try it again.

[MontanaRails]

I didnt see any effect either. One thing I did notice is that AI trains do not seem to abide by changes to the coupler configurations. I'm sure we all know how goofy and stretchy the AI trains get in the "Observer" free roam in the HSC route. I made significant changes to the couplers from posts in the forum (which, as a player driven train, work GREAT!) but as an AI train, they act exactly the same way as before. This tells me the scenario must code something about the actual shape and position of the train, so changes to the dynamics, after its written, have no effect on it. I guess it would have to be re-scripted with the new coupler dynamics.