HSC F7's Dynamic Brakes

[Railfan587]

Does anyone know why the HSC F7's dynamic brakes are so bad? I've been trying to finish the HSC scenario "People Power" and when come through the tunnel and start my descent, I turn on the dynamics. But, the dynamics are so bad that don't control the train's speed well or they make the train stop completely or glitch out the train and makes it derail.

[buzz456]

I suggest since you posted on another thread about dynamic brakes not working properly that perhaps you should think about your technique. Many locomotives won't even switch to dynamic brakes without a ten second delay, further changing the settings very gently seem to be most effective. To get this feature to work properly you really need to anticipate the need. I find that one of the most fun things in RW not something that doesn't work right. Forgetting about the percentages it's probably one of the more realistic effects.

[Kali]

Most stock from that era came with dynamic brakes set up with the wrong units, so they're ( I think ) about 4.5x too strong - added to the squashy couplers and you get your train collapsing ( there was a lot of work done when HSC came out about this if you give it a search ). Have to be *very* careful or edit the dynamic brake files.

[Railfan587]

Most stock from that era came with dynamic brakes set up with the wrong units, so they're ( I think ) about 4.5x too strong - added to the squashy couplers and you get your train collapsing ( there was a lot of work done when HSC came out about this if you give it a search ). Have to be *very* careful or edit the dynamic brake files.

That is exactly what I'm talking about. I do know about the 10 second delay with the brakes and that you have to be easy with them, but braking on that particular F7 dosen't work well. The default Kuju F7 from TS2012 has much better and smoother brakes brakes. Maybe I'll change the brake simulation to the Kuju F7 brakes.

[Ericmopar]

It's the brake settings they are way wrong. I have that one and may have fixed it. I'll have to look at the bins for that engine.

It really is wrong Buzz. They jam on even at minimum settings. The air brakes are none to effective either on some of the locos for that route. Off or on it seems, with not much in between.
I haven't driven that for awhile so I'll have to take a looksy.

[Railfan587]

I switch out the HSC F7 simulation for the Kuju F7 simulation so I could finish the scenario. I made it down without a drive quality or speeding penalty, but then got hit with a HUGE penalty for being 4 minutes late to Altoona and lost more than half of my points . I didn't even get a bronze. I think the game glitched at Cresson because it keep me waiting more than a minute after the red loading bar was full.

[MontanaRails]

To fix the dynamic brakes, you need to go here: RSC\HorseshoeCurve\RailVehicles\Diesel\F7\Default\Simulation and open F7A_Simulation.bin (with RWTools or similar).

Search for "<DynamicBrakesSystem>" and change the <MaxForce> value to around "60" or "70". I used 67. The default should read "267". It's WAY too high.

f7.JPG

Hope this helps.

[Ericmopar]

I switch out the HSC F7 simulation for the Kuju F7 simulation so I could finish the scenario. I made it down without a drive quality or speeding penalty, but then got hit with a HUGE penalty for being 4 minutes late to Altoona and lost more than half of my points . I didn't even get a bronze. I think the game glitched at Cresson because it keep me waiting more than a minute after the red loading bar was full.

I've been looking at the bins for the cars and engines in HSC. There are a few things wrong with the engines and car bins.

One of the Geeps had the maximum force for the dynamics at over 260 when it should be closer to 30. The F7 was over 100...
Things like tripple valves in the P70 coaches are set to eFalse when they should be eTrue.
The settings are all over the place, but I'm going to go thru them tonight.
I'll post something up tonight or tomorrow.
That maximum effective speed isn't correct either.

[Ericmopar]

You guys with Horseshoe Curve can try this. I had already made mods to most of it.

I went through all the stock on this one and test drove it, so it should be good to go.
It may overwrite certain engine mods, if you already have some in the engine and/or car simulation bins.
Please read the Read Me first.

HorseShoeCurveBrakeMods.zip

[Kali]

The thing I think uses SD40 data - this was and probably still is a common thing for RSC, there's UK stock that is just reshelled versions of other UK stock. Somewhere or other in here there is a much fixed up set of sim data, but it's probably from 2011... There is a chance it's been given kN numbers and is reading them as kilo lb-ft - it would be the only thing that actually reads lb-ft even if other things say they are but really that would not be a surprise - but on the other hand it might just be too big full stop, like someone has done a unit conversion when it wasn't necessary. F7 continuous TE is in the order of 175kN so it could never be bigger than that even if the dynamic brake system worked as well as the power generation.

Anyway this thread might be a useful read: viewtopic.php?f=6&t=3737 ( and I found the coupler one viewtopic.php?f=29&t=4040 )

Beware of randomly ticking boxes: I haven't actually looked at the P70s but if they have a distributor that's set up to work like a triple valve ( might be a workaround for another brake issue ) then no they won't want the triple valve box set to "true". The dynamic brake is not just the MaxForce entry but the graph too, which is a cap to the game's generated dynamic force number. I would hope the dynabrakes aren't as odd as practically everything bar straight electrics, but you might find MaxForce on it's own doesn't do what you expect...

Edit: reading back over old threads I seem to have been using 45kN for a SD40-2, so 60 is a bit much for a F7 still...

[Ericmopar]

I set up and test drove everything. Those are correct settings in my mods, including the couplers.

These guys keep setting up ultra stiff spring coefficients, no minimum distance etc out of habit, but I found out the settings work best when a person sets up a minimum distance of .01 and a maximum of .03 to .05 depending on whether it's freight or passenger stock.

The damping is something that third party couplers need, but not RSC's. I usually add some to smooth out the ride, but real couplers on real American freight trains, have hard stops and do shake up the crew in the cab, when slack runs in and out.

Spring coefficient is a no, no. Thats the big one that creates springy, spongy train effect. People that use spring coefficient, have to use huge amounts of damping as well.

Target distance should also be at 0. That one causes creeping train after a stop effect, and has nothing to do with centering the coupler like some people think. It's a setting for buffer style couplers on European stock. As are spring coefficient and damping.

I know from the comments I get around the forum, that no one except Mike Rennie has even tried my settings with the RSC couplers. They just put the wrong settings in without even trying mine first.

I've found RSC's own Wiki to be frequently wrong, and claim settings don't even function when they do in fact still work in game...

I think it was Buzz that told me where to start with dynamic settings. I found his settings to be spot on for the different generations of diesels.

That forum info from several years ago is quite frequently outdated now.

[Kali]

0 spring couplers work fine for anyone, Michael Stephens are all set up like that - as they should, there's no special RSC coupler blueprint. The high spring force thing was an attempt to work out what it should actually be given the physics seems to use Hooke's (spring) law and the bp figure seems to be in N/m. Then we ( might have been you Eric? ) found 0 entry worked as distance clamping didn't work so we needed impossibly high spring values anyway. Having a giant number of 9s until it hits the biggest number you can store will effectively do the same thing, just looks horrible :p

As I was curious I did some experiments with mid-point couplers, as the buckeyes are. Locomotive used is a GP9 ( so fairly heavy relative to the test boxcar ) with rigid spring couplers with 0.05 MaxDistance ( not enough to matter ), the experiment car has no rolling resistance for now.

--

First: I set MinDistance 1m, MaxDistance 2m, TargetDistance 1.5m, receiving point 1m ahead of the body pivot just for the sake of it ( given it's explicitly stated that it does nothing for mid-point couplers ) and unfortunately the test coupler didn't couple until the pivots overlapped. At that point the car violently sprung away as you'd expect given they're trying to move to min distance instantaneously ( given we have an absurdly strong spring with spring at 0 no surprises there ), sufficiently violent to derail & fall over.

=== Conclusion: MinDistance is not added to body pivots until after coupling ( facepalm, unsurprised though ), no receiving points for mid-point couplers as advertised. Coupling will immediately try and make the car move to collision pivot + MinDistance using the coupler spring force.

To get a better idea what's going on, now I set the spring coeff to 200 which is enough to move the car, but not enough to throw it off the track like the previous. Damping still 0. The car didn't couple until the body pivots met again, this time it bounced away & yoyo'd for a while until I put it's handbrake on. I pulled the coupler out until the car moved slightly & then stopped, then let the handbrake off. No movement. Pushed the coupler to it's minimum distance - this did actually work now it's coupled - stopped again, put the brakes on until everything settled and then let them off. Result of that, car very slowly creeping away, so a bit inconclusive. At that point I left the handbrake on & attempted to shove it with the locomotive, and then let the power off - was enough to spring the locomotive off relatively hard.

=== Conclusion: not sure TargetDistance actually does anything.

So this time I set TargetDistance to 1 as well. Set the handbrake on the car, pulled the couplers out to 2m, let off the handbrake, car stayed put.

=== Conclusion: TargetDistance apparently does nothing for these couplers.

Next, giant numbers in spring coeff, just to check. Put in about 30 9s ( which won't fill a 64bit number but is way more than enough ), car behaved like a 0 entry.

=== Conclusion: 0 entry for spring coeff probably just sets it to maximum ( from a mech eng point of view, definitely ).

Damping: Spring coeff 100, damping 100, target 1m. Coupled, car rolled away at constant speed until max distance, and then slowly returned to min. Put the handbrake on the car and pulled the coupler out to max distance, stopped, gave a brief burst of power ( about 0.3mph ), put the loco brakes on and let the handbrake off. Car rolled at constant speed until min distance, and crept back - so it's being fully damped in the return direction. Target distance still doesn't seem to be doing anything.

=== Conclusion: *scratch head*. New numbers!.

Damping 10 this time. Coupled ( note I'm coupling with the independent brake on so the engine doesn't move ), car sprang away and oscillated back and forth, reducing speed at each change of direction, and as expected 10 times.

=== Conclusion: coupler slack is getting in the way.

Max Distance is now the same as Min distance at 1m. The car still oscillated back ~10 times ( hard to see the last because they're so small ). Just a confirmation of the above, so then I set the damping back to 100, which stopped the car at 1m as expected.

=== Conclusion: Damping seems unsurprising.

Damping now 1000, to see if it overpowers the coupler spring.

=== Conclusion: No, coupler still returns to "neutral", but apparently ( given I didn't time it ) 1/10th the speed. Also unsurprising.

Damping now 0: Car oscillates for quite some time ( not actually indefinitely, so I assume there is some other thing getting in the way, but it's very small ).

=== Conclusion: Damping value of 0 actually means approximately 0.

One last test - MinDistance to -1m. Coupled at the usual distance when the pivots met, and then the locomotive carried on *inside* the car.

=== Conclusion: couplers take over inter-car collision detection on coupling, in some form. The hit boxes eventually seem to collide, so they do actually do something outside the editor...

For the sake of it again I tried MinDistance 1m and MaxDistance 0m, in case the clamping code was just backwards, but that didn't seem to change anything.
--

Overall: no extension clamping means the springs are still practically useless. Given I didn't try 100 cars here I've nothing much to say about damping ( some may help stop jitters and a lot will probably cause creep, but that's speculation and another test ). Coupling happens via the collision detection ( vehicle pivots ), once they're coupled then the couplers take over. MinDistance other than 0 forces the couplers to force the vehicles to pivot+MinDistance on coupling - with max spring force couplers that may be quite violent, especially if the game starts physics for a complete train at once. MaxDistance works as expected, so slack is Max-Min. TargetDistance I couldn't get to show any effect.

I'm never sure I didn't miss something obvious here - as this is just an experiment report, fully open to criticism of it. If you like your own bunch of numbers, this isn't telling you to do anything else...

--

I worked out why I used a 45 dynamic brake figure for the SD40, it was to account for the gearing - that gives 45,000lb-ft dynamic brake max, which looks good for a higher geared one ( I suspect I actually used a SD40 manual and not a -2, mind ). BP figure does indeed seem to be kilo-lb-ft then as people said, hurray for consistency.

[Ericmopar]

So.... you didn't even try my coupler mod for HSC without changing things first?

All this scientific mumbo jumbo, trying to analyze something that I know works like a charm.

I'm starting to realize that some persons around here, must not be trying my stuff out without hacking it first.


I've got all my American DLCs couplers set to those specs, except the RW&A stock. It doesn't like those settings to much. For some reason on the RW&A, those coupler settings don't take effect at all.